A senior IT executive at a major pharmaceutical company summed up the challenge for Linux at the ZDNet UK IT Priorities conference when he asked one simple question: what are the benefits in migrating from Microsoft to Linux at the desktop?
Linux on desktop will be ‘cheaper and more secure’, claim vendors
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
this article really nails a lot of great linux benefits:
1) great multiuser support. 1 linux box can easily handle several simultaneous users (i read an article earlier this week about a setup that included 10 monitors/keybords/mice connected a single box (not thin clients) where all could work simultaneously, unfortunately i can’t find the article)
2) application selection. don’t want a web browser or media player installed? you don’t need to install ’em. how many mall kiosk workers, airline agents, or cab drivers need a web browser or a media player on their computer? probably close to none. since windows lite won’t be available in the US, companies will be stuck offering these productivity vacuums to their employees as long as they are embedded in windows. there is nothing remotely like this in linux. don’t want a web browser or media player? don’t install one.
3) price – windows server enterprise is $4000, plus around $40 for each client. for that guys company with 20,000 employees, that would be well over $800,000 in windows licenses (there’s got a be a cheaper site license, right??) cost of enterprise linux, plus unlimited clients: 0. if windows licenses do cost around $800,000 (which i’m skeptical about my math), you could hire 16 new admins with the money you just saved, though you wouldn’t need them because linux administration is a breeze.
4) administration – for the most part, linux administration comes down to editing text files. if your admin can’t edit text files and read man pages and other documentation, i wouldn’t want him working on any servers, let alone servers that try to force your hand with GUI tools or make you edit a cryptic registry file. any admin (windows, linux, unix, or otherwise) worth his salt should be able to admin a linux box in under a month.
so not only is your hardware supporting more users, but you’ve removed possible distractions from your employees while saving money and helping your admins learn something. everybody wins!
I came to Linux for the soul reason of IT’S NOT MICROSOFT, and found out along the way that there were many more benefits.
This article does indeed point out some great benefits, but my favourite still remains… IT’S NOT MICROSOFT and I’M LOVING IT!
Applications dictate what OS goes into most shops. In many existing shops it can’t be changed due to proprietary applications currently in use. In other shops the cost of MS licenses can dwarf the cost of migrating existing proprietary applications to a run on a non-windows server/desktop.
Pointing to only the cost of licensing for MS products is often minor when looking at the overall picture.
Pointing to only the cost of licensing for MS products is often minor when looking at the overall picture.
Not if both systems meet the requirements. Even if you include the extra staff education, going with linux will most likely be the best and cheapest solution.
For the typical office, schools etc. linux will do just fine. Not all companies run certain enterprise apps. Actually, the percentage is probably pretty low.
I just don’t understand why people are so obsessed with one OS ruling the world. Haven’t we seen what monopolies does to the industry, and consumers allready? No matter if it’s opensource or not, choice is good IMO. And I’m not talking about different linux distros.
I’m hoping for a world where windows, osx, linux, skyos, haiku, syllable etc. all have decent third party support. Where a company decides to support a platform, not because it’s the most popular one, but because it’s the best one.
Well, one can dream right.
I just don’t understand why people are so obsessed with one OS ruling the world. Haven’t we seen what monopolies does to the industry, and consumers allready? No matter if it’s opensource or not, choice is good IMO. And I’m not talking about different linux distros.
User: Hey, where are all the applications?!?
IT: Look up, now to your left, you see that applications menu?
User: Well who’s stupid idea was that, they belong under start!
I blame it on the schools, we rely on the familiar because we’ve forgotten how to think. Actually, if anything is to blame, it’s probably time.
I’m hoping for a world where windows, osx, linux, skyos, haiku, syllable etc. all have decent third party support. Where a company decides to support a platform, not because it’s the most popular one, but because it’s the best one.
Unfortunately the large businesses, which appear to be on the front line right now, are driven by the same thing as Microsoft: greed. It’s all about money, and until it’s proven without a doubt that Linux is cheaper (and perhaps it really isn’t always cheaper, at the moment), then they won’t switch.
Well, one can dream right.
It’s hard to trust someone who doesn’t.
This can be a serious issue! Most of my development work over the last ten years has been with the Canadian Federal government and I can tell you that large organizations have an enormous amount of custom written Visual Basic apps, and IE specific web applications that would need to work seemlessly on a Linux desktop. Any large organization would need to independently verify that each app would work on the proposed system before rolling it out. Running an App in VMWare or with WINE/Crossover office would still require the testing effort. Imagine that cost! It takes them years just to roll out new versions of windows. What is the solution to this? It would be nice if all applications were written in cross-platform a language or toolkit, or if web applications supported other browsers, but that isn’t the case.
“The open-source methodology produces better code as more are people looking at the code,” said Jollans.
Well, just because MS writes shitty code doesn’t prove this guy’s point
Take Opera vs Firefox for instance .. which one is more well written?
He conceded that some customers might be better off with Linux, but said that customers should make a case by case evaluation based on their own requirements.
This is a great point. Find out what you need to do, look at the kind of applications that each OS provides, and then make your choice depending on the applications, not the OS. I’ve seen a lot of companies playing politics and saying ‘we have to use this application or this technology’, when it really wasn’t the best option for them.
“Sometimes Windows costs more than Linux, sometimes it’s the other way around. Customers need to look at their operating environment,” said McGrath.
This is also a good point. If you’ve got a $500 professional app and also a free app that does everything you need, then use the free app. However, if the free app is “Joe’s Open Source App” that’s currently in alpha v0.2 and contains more bugs than a roach motel, you need to consider your options carefully. I’m sorry, but open source doesn’t always mean better.
“Our customers don’t want employees to play Solitaire all day. Novell has invested heavily in kiosk technology where the desktop functionality is cut down, so that it only includes applications which companies want their employees to use,” said Schlaeger.
Obviously, some people have never heard of nLite.
Microsoft’s McGrath denied that Linux has any advantage over Windows in this area and said that since Windows 98, customers have had the ability to reduce functionality in the desktop using profiles and policies.
Also true. The IT department where I work has Win2k locked down so tight, I can’t even view the damn calendar when I double click on the taskbbar (highly annoying).
“Who guarantees that you can access your information from a proprietary file format five years down the road? Open standards relieve you of that pain. I use an open-standard file format so that I can keep the keys to my information,” said Schlaeger.
Not only is this a flawed argument (if you’re using Joe’s Open Source app and he gets bored developing it and quits in a few months, what are you supposed to do with your data), it’s really not a case for Linux, so open file formats DO exist for other opearing systems as well.
“Not if both systems meet the requirements. Even if you include the extra staff education, going with linux will most likely be the best and cheapest solution.
For the typical office, schools etc. linux will do just fine. Not all companies run certain enterprise apps. Actually, the percentage is probably pretty low.”
I can tell in a heartbeat you have never done this on a large scale. Systems are what systems are. You don’t get to make choices here. You live with what you have, and you don’t proceed to dick with it; because you CAN.
You can’t just “do it” in corps that have been using Windows for a decade. Too many choices have already been made, software written, and an entire method of doing business being conducted.
Hell I suppose you could… Ya might as well file for bankrupcy at the sametime.
So how do you demonstrate to them that this lock in is in fact a problem, and that they can eventually fix it if they start the transition to a more portable OS? Do they generally feel the constraint, or do they not really care as long as things are working?
I want to see greater GUI work for the desktop, that means better thought process on GUI design, HIGs and philosphy. I also want to see alot more work going into pushing updates out quickly and efficiently for security and future release purposes.
With all due respect,
Here are some observations about your comment (please do not take offense).
1) “custom written Visual Basic apps”
2) “IE specific web applications”
possible soloutions posed:
1) “Running an App in VMWare or with WINE/Crossover office would still require the testing effort.”
2) “Imagine that cost! It takes them years just to roll out new versions of windows. ”
“What is the solution to this?”
Dont code in a proprietary language. I know its a little late now. However, a basic rule, if you use custom app’s made by custom programming launguages, you asking for trouble. It was and is the job of the programmers to make applications portable (if requested).
Disclaimer: I am not a programmer, however, I have to deal with various cross platform issues in networking. Does product A by vendor A interoperate with product B with vendor B. Now granted, each vendor has nice little toys to make the network folks job easier. Now by going down that path with proprietary tools and toys makes adding new equipment quite difficult (IE network management software). By going with standard protocols and not using custome vendor tools, adding new technology becomes a tad bit easier. Your not tied into 1 person/company. Vendor A gives you a difficult time, switch to vendor b or c or d. There is a lot to be said for vendor lock-in and not being locked-in. Most companies and people dont consider lock-in, until it happens to themselves. Once your in, you in and its not pretty trying to get out of the whole that was dug. For the last 10 years, I have continually tried to avoid being to any one company and to a good extent, it has worked. What to do after the fact? Sorry, cannot help you there. Modular is better than custom, any day of the week (IMHO). Replacement parts are cheaper and easier to obtain (U.S. Industrial age humor).
“It would be nice if all applications were written in cross-platform a language or toolkit, or if web applications supported other browsers, but that isn’t the case. ”
There were cross platform languages, it was the programmers/companies/gov choice not to use them.
Once again, vendor lock-in, no matter where it comes from or what industry, show be avoided at all costs. What happens if your vendor goes out of business or refuses to support product X at this set date and time.
I work at a big pharmaceutical company. The desktop OS is largely not the biggest concern for us. It’s the apps.
Most pharma companies use CRM apps such as Siebel which have no native Linux clients (and they don’t plan to). My company and my ex-company, Novartis, both use Notes as an exchange platform. Hoffman-La Roche, Aventis/Sanofi all use exchange/outlook exclusively for “all their global offices”.
This includes our mail, training, expenses and a host of other apps (corporate policies etc). None of these can run on Linux. IBM may tout the fact that they are linux friendly, but yet they’ve failed to come up with a native port of Linux notes. Most companies, including ours, takes prescription data from another company called IMS (and now Brogan). This data is sorted into very complex excel spreadsheets and distributed to sales force/management for analysis. All of this information cannot be ported to Linux easily. Nothing, not even staroffice/OpenOffice.org, can do what Excel can and porting all of these very complex spreadsheets (monthly IMS data, territorial/managerial cost center expense reports/spreadsheets, monthly sales spreadsheets etc.) do not work with any OSS solution out there.
In addition, the biggest desktop users of a pharma company are the sales force. 99% of them (trust me, I know) have never even heard about Linux. I’m not sure if you can comprehend this, but the costs involved in training them are outrageous. In addition, the ROI you get from moving from Windows to Linux is quite meager (and not Windows Server doesnt’ count as this is desktop realm).
So to sum up, lack of support from big vendors such as Siebel, IBM, SAP (in our case) prohibit any plans to move to Linux. In addition, there is no ROI from simply moving to Linux from Windows as this (the OS) is a “very” small part of the big picture
“So to sum up, lack of support from big vendors such as Siebel, IBM, SAP (in our case) prohibit any plans to move to Linux. In addition, there is no ROI from simply moving to Linux from Windows as this (the OS) is a “very” small part of the big picture”
BINGO! Who gives a damn about the OS when the vendors you are dealing with couldn’t careless themselves.
Unless your vendors are going to float the boat for your migration; you are screwed before a single shot was fired.
Been there, done that.
This article is an opinion coming from the people involved with Linux migration projects. It’s impossible to argue with some ones educated opinion for its well an opinion.
Typical FUD.
It turns out that SAP, Siebel and IBM are actively porting their software to Linux. In fact , much of Siebel’s application stack already runs on Linux.
Lotus runs on Linux via Wine and does so reliably. Additionally, IBM is now java-enabling Lotus so that it will run everywhere.
“Typical FUD.
It turns out that SAP, Siebel and IBM are actively porting their software to Linux. In fact , much of Siebel’s application stack already runs on Linux.
Lotus runs on Linux via Wine and does so reliably. Additionally, IBM is now java-enabling Lotus so that it will run everywhere.”
Not at all FUD.
Who wants to run anything “via” something in a production enviro? Nobody. Lotus isn’t a hot ticket less java, can’t wait for it to get more hosed so it will run everywhere.
“Who guarantees that you can access your information from a proprietary file format five years down the road? Open standards relieve you of that pain. I use an open-standard file format so that I can keep the keys to my information,” said Schlaeger.
“Not only is this a flawed argument (if you’re using Joe’s Open Source app and he gets bored developing it and quits in a few months, what are you supposed to do with your data), it’s really not a case for Linux, so open file formats DO exist for other opearing systems as well.”
It is not a flawed argument. Not all proprietory software sticks to open standards because you can’t lock people in that way.
If you are using something proprietory, and they go bust – where do you go?
If Joe gets bored developing his open source app, at least you have access to the source code to continue with it. Most open source software I’ve seen and used usually uses open standards so the data can be imported easily.
In theory, it might (and that’s a big might) be possible to use the Visual Basic to VB.NET source code translators that VS.NET provides (I’m pretty sure there’s some tool out there that does it), and then re-compile on Mono once their winforms implementation matures. Of course there will be problems, but crossover could be a temporary solution until the developers get things worked out. At least they’re not rewriting all of their code.
Yeah good old lock-in. Greed. Anti-Competitive practises. You name it, the computer industry has them. So who does one trust in such an industry where truth hath many faces, and where monopolies do what they like.
Where will Apple be 10 years from now. Where will Linux be in 10 years time. Where will Microsoft be in 10 years.
There are people who will say that any one of these will be dead.
Myself I can’t see either Unix or Apple dead in that time span.
CLUELESS…. thats all I can say.
IBM “Linux Friendly” ?? because they did not release a native copy of Lotus Notes ? hmm, why spend the development time on something that works flawlessly through WINE ?
IBM are in fact one of the biggest Linux supporters there is.
Now you also tried to raise a point about apps, however, you inadvertently gave Siebel as an example. I also use Siebel, but I run it NATIVELY on a Mandrake system. It is the front end to the server which is running Siebel NATIVELY on a Debian server, hmmmmmmmmm how could that be ?
I will tell you how.. Siebel is a JAVA app.
Well Siebel version 7 is, thats what we got here.
See, the main point you should have raised was this.
The 99% of people who do not know what Linux is, should be made aware of it.
You can’t just “do it” in corps that have been using Windows for a decade. Too many choices have already been made, software written, and an entire method of doing business being conducted.
I know that it’s not something you do during a coffebreak. But it is in my opinion something that’s worth investigating. It can actually be worth it.
Besides a lot of enterprise software is written in java and is most likely to run on linux. So that’s not a problem in every case.
It is a big decision to make, and it should be considered many times. But the important part is that it’s considered.
You’d be pretty stupid if you didn’t investigate options that could save the organisation both money and time, even gain productivity. (I’m not saying it will, but it can)
Also, the small companies counts too, because there’s a lot of them, it can mean a lot to the market if they switch.
I’m not a big fan of linux on the desktop myself, even though I use it on a daily basis. But I do see it as a serious option, not just a “cheap alternative”. It’s a much more sound alternative than Mac IMO, even though I like OSX more.
Not at all FUD.
Who wants to run anything “via” something in a production enviro? Nobody. Lotus isn’t a hot ticket less java, can’t wait for it to get more hosed so it will run everywhere.
I don’t think anybody sees WINE as an end-all solution. That’s not what it’s meant to be. However, it does provide a way for people to migrate from Windows who have invested thousands or millions in their programs. So then they will have the choice to a) keep it how it is or b) port it to a native application over time.
Its going to take more than this article to convince me to switch to using linux on my desktop. Especially in a business setting. I won’t switch just because its the cool thing to do nowadays.
“security costs can add between 10 and 20 percent”
Doesn’t mean it has to.
“Chris Schlaeger, the vice president of research and development at Novell, agreed that cost was a big benefit.”
Hrmm… what would you expect a competitor to say?
“”The open-source methodology produces better code as more are people looking at the code,” said Jollans.”
Wrong again. I don’t know anyone who just sifts through the code. Its only the developers of said programs doing the looking, not the end users of linux.
“”Our customers don’t want employees to play Solitaire all day. Novell has invested heavily in kiosk technology where the desktop functionality is cut down, so that it only includes applications which companies want their employees to use,” said Schlaeger.”
Then don’t install it. Use local and group policies to restrict access. Sheesh.
I’ve already pointed out 4 things wrong with the article and their flawed arguements. No point in adding more.
I am not sure what the Canadian govt is doing but I know from my own experience in working with, and in, the US gov’t is that you use 3 things. 1)Office Suite. Any will do the job. 2) email client; again, any will do. 3) The other programs we use were all java based, that way they were accessible to anyone and everyone. I even knew guys who logged into a remote Unix machine and ran their desktop from there. Sure there are some proprietary apps out there that only work in Windows. But for the rest. Linux will be cheaper.
Ever tried planmaker (spreadsheet)?
You are 100% correct. AutoCAD, SolidWorks, Inventor, SolidEdge, IronCAD, etc along with most other high end CAD (2D and 3D parametric modelers) only run on Windows. You may find a few extremely high end CAD apps that run also still run on some proprietary form of Unix, but are either unsupported on Linux or just won’t run on Linux.
Pro/E will run under Linux and is supported, but now you’re talking about a $10,000+ investement just for the application alone.
I’m sorry Linux zealots, but purchasing, supporting and maintaing these types of applications drawfs the very small cost of Windows, including administration. For the price of 1 seat of AutoCAD I can buy 26 seats of Windows XP Pro. Or let’s talk about the $5,000 workstation I need to run SolidWorks, which itself is another $5,000. Or would you like to talk about FEA packages?
I like Linux, I use it from time to time. But if I can’t have a good quality CAD application for a decent cost on Linux (sorry Pro/E and UG are just too dang expensive) I can’t move off of Windows 100%. And no, running an application via WINE in a production environment doesn’t cut it. Native ports only, IMO.
Do Original software vendors support their applications when run through WINE? if not, then I would have thought production environments are off-limits to WINE and its ilk. IMHO WINE is very good as a stop-gap method for getting something working while looking for a permanent solution, but is not a solution in itself.
In my opinion the best approach for a large organization migrating to a Linux desktop which has problems with a limited number of Windows only apps, that are used by a small number of specialized users, is Rdesktop. Run a few Windows servers with these applications and serve them remotely using Windows terminal services to Rdesktop clients on the Linux desktops.
The solution to the Notes problem is to upgrade to Lotus Workplace 2 which is a cross platform Java application based on the Eclipse platform and intended as Notes successor.
This makes what, the 1 millionth on osnews?
In the pharmaceutical company I work for migration to a Linux desktop would be be a lot easier now than it used to be. Our ERP runs on Oracle (on AIX and Linux servers) and of course the intranet front end works with Mozilla as well as IE.
In the laboratory our Chromatography management and LIMS systems both run on Windows. However both are now used from Citrix servers. I have been able to use them from the Citrix Linux client on my Linux box without problems.
What stands in the way is the Windows entenched ideas of those parts of the IT department managing the corporations desktops.
CLUELESS…. thats all I can say.
Heh…surely coming from a non-pharma person, this is quite cute.
IBM “Linux Friendly” ?? because they did not release a native copy of Lotus Notes ? hmm, why spend the development time on something that works flawlessly through WINE ?
Right.. And the company will pump in money to move to Linux and run wine but still get a win32 apps to run via wine. Not mention, the costs associated with training the sales staff. Sure, mate…makes sense
Now you also tried to raise a point about apps, however, you inadvertently gave Siebel as an example. I also use Siebel, but I run it NATIVELY on a Mandrake system. It is the front end to the server which is running Siebel NATIVELY on a Debian server, hmmmmmmmmm how could that be ?
Siebel SFA is not a java app. The server can run on most unices (we run it on Solaris) but the clients (SFA clients for desktop use for most sales forces “do not” run on Linux). In addition, are you using their pharmaceutical industry product? You do realize that Seibel is customized to each industry, company and that each company has to pay millions for this?
Not to mention, bigger pharma companies, such as Novartis, still run Windows NT Workstations with a custom win32 app for CRM which backends to their Oracle run servers.
I have noticed a number of people saying Lotus Notes runs well under Wine. I would be interested to hear from those who have experience of doing this. The only information I have so far is one article claiming that it worked, but not perfectly. The thing is that it would need to work perfectly (no exceptions) because companies that use Notes tend not to be able to work without it.
Looks like Lotus Notes works pretty darn well under Wine to me:
http://appdb.winehq.org/appview.php?appId=27
With all due respect,
Umm…okay.
“What is the solution to this?”
Dont code in a proprietary language. I know its a little late now. However, a basic rule, if you use custom app’s made by custom programming launguages, you asking for trouble. It was and is the job of the programmers to make applications portable (if requested).
You offer a solution and then admit that it is not really a solution because it’s a little late now. That was my point!!
“It would be nice if all applications were written in cross-platform a language or toolkit, or if web applications supported other browsers, but that isn’t the case. ”
There were cross platform languages, it was the programmers/companies/gov choice not to use them
You have completely missed the point of my comment. Your solution is to pontificate about why the problem exists. Thanks, that’s really helpful.
All applications, no matter how small, would need to be thoroughly tested on the proposed platform. That would take a large amount of time and money. What about existing enterprise software that doesn’t exist on Linux (Workflow, Document Management, Electronic Forms, ERP systems, Records Management Systems, etc, etc, etc)? I know of government departments that took 18 months to upgrade from Netscape to IE.
A solution for a slow migration:
1) Force all future development to be done with a cross-platform language or toolkit (Java, QT, Python, etc)
2) Web apps need to work on all browsers
3) Convince all of your enterprise software providers to support other platforms.
Items 1 and 2 would take time but could be done since the organization controls this. But there would be a training cost for the internal development staff. Testing and QA costs would also increase for each application. Good luck with item 3 especially if you’re the only customer making the demand.
Migrating to Linux on the desktop for any large organization isn’t simply a matter of downloading an ISO image and installing it on everyone’s PC. It’s the cost of migrating, testing, and supporting the existing application base on the new platform that will be the real expense.
These costs and the associated risk with users not happy with how their applications work (or don’t) on the new platform, would surely convince most CIO/CFO/CEOs that staying on the MS upgrade cycle is worth the pain and expense. That doesn’t make me happy, but that is reality.
P.s.
You’re != your
“1)Office Suite. Any will do the job.”
It sure will. Right up to the point where you are being sent a large volume of files in non-supported formats. Then the whole idea falls apart.
All of the office suite create documents well, the interchange between those same suites is another matter. You can generate docs in XML, PDF, and RTF, and do these exchanges. Someone planning to tell your best customer in Chicago to follow these formats?
You can’t dictate the format of the documents and files you recieve from your customers. Unfortunate as it may seem Microsoft applications largely dictate the formats used by business at present. There are exceptions to everything, but Word sets on most business desktops.
“Migrating to Linux on the desktop for any large organization isn’t simply a matter of downloading an ISO image and installing it on everyone’s PC. It’s the cost of migrating, testing, and supporting the existing application base on the new platform that will be the real expense.”
Amazing how simple the truth is.
In theory, it might (and that’s a big might) be possible to use the Visual Basic to VB.NET source code translators that VS.NET provides (I’m pretty sure there’s some tool out there that does it), and then re-compile on Mono once their winforms implementation matures.
In practice, such migration can be big pain in the ass. I’m working in a little company, developing one
app for years, last years we’re using VB6. Of course we’ve experimenting with VB.NET and code translators – it [VB.NET] is highly incompatible with all little and big tweaks, included into our main VB code. These tweaks, btw, are one of the most important part of our app’s speed, GUI, visual look etc – all what the users love (and pay for). If even .NET is incompatible with VB6, how about mono and .NET compatibilty then? I think it’s somewhere about 80% or less (but I hope I’m wrong).
Some years ago we weighted possibility to use java (cross-platform!) – no way. Interface looks ugly, is slow, no good components for data manipulating and reporting (well, things are somewhat changed since then) – and no users, needing something other than winodws.
Truth is that there’s no point make commercial software for linux, at least not for little companies.
About WINE – so far no success to make our app work with it. Installer uses COM “in not supported way”. Component installers use COM “in not supported way” or just crash. Copied executable (+ all dependent files) won’t run – similar and other errors. And we have no incentive to develop WINE further – nobody will pay us for this.
1) great multiuser support. 1 linux box can easily handle several simultaneous users (i read an article earlier this week about a setup that included 10 monitors/keybords/mice connected a single box (not thin clients) where all could work simultaneously, unfortunately i can’t find the article)
So can Windows. I’ve heard of keyboard/monitor/mouse sharing applications for PCs going as far back as DOS. They’ve never been particularly common or popular for things other than, say, cash registers and the like.
2) application selection. don’t want a web browser or media player installed? you don’t need to install ’em. how many mall kiosk workers, airline agents, or cab drivers need a web browser or a media player on their computer? probably close to none. since windows lite won’t be available in the US, companies will be stuck offering these productivity vacuums to their employees as long as they are embedded in windows. there is nothing remotely like this in linux. don’t want a web browser or media player? don’t install one.
It is not hard to remove these applications from a Windows install if you really want to. In particular, if you can expend the effort to do a custom Linux install, you can expend to effort to do the same thing on Windows.
3) price – windows server enterprise is $4000, plus around $40 for each client. for that guys company with 20,000 employees, that would be well over $800,000 in windows licenses (there’s got a be a cheaper site license, right??)
Yes. No company with 20,000 employees is paying anything close to the retail price. Heck, any company with more than a hundred-odd employees should be able to start negotiating for some level of discount.
cost of enterprise linux, plus unlimited clients: 0.
20,000 employee enterprises do no run on free versions of Linux downloaded off the net. The cost of “enterprise linux” is vastly higher than $0.
You also ignore the simple fact that software licensing is usually one of the least significant costs incurred on IT infrastructure. $40 per user ? That’s probably less than an employee will cost a company in an hour. Average that cost out over the ~6000 hours for the minimum 3 years that $40 will last for and it’s practically insignificant.
4) administration – for the most part, linux administration comes down to editing text files.
Correct. The value of the unix administrator comes significantly from his ability to remember the dozens of different configuration file formats and their locations in different versions of each *nix OS.
if your admin can’t edit text files and read man pages and other documentation, i wouldn’t want him working on any servers, let alone servers that try to force your hand with GUI tools or make you edit a cryptic registry file.
HTF is using GUI admin tools “forcing your hand” ? How is editing “cryptic registry files” (a very rare occurrence adminning Windows) any different to editing cryptic text files (a very common occurrence adminning unix) ?
Also, one big advantage of GUI admin tools over hand-editing text files is that they provide input validation.
so not only is your hardware supporting more users, but you’ve removed possible distractions from your employees while saving money and helping your admins learn something. everybody wins!
You have not demonstrated how any of the above would actually occur in reality.