Moving Small Organizations from Windows to Linux?

chris1646 asks: "Currently we are a small organization that is entirely a Windows shop. Next year much of the server and desktop hardware we run will need replacing. I am looking for creative ways to introduce Linux as my desktop and server OS of choice, however a couple of our core applications run exclusively on Windows. Has anyone had any success hosting Windows applications via terminal server while using Linux as the client OS? Has anyone handled a AD to open source LDAP migration?"

Comments

Look at costs, Servers first

[innosent]

Look at your costs before migrating to Linux clients for RDP. Terminal Services Licenses cost nearly as much as a full XP license, so you will likely spend more to do it this way. Having said that, you might be able to run your critical apps under WINE, and use Linux X clients to run it via SSH. I would definitely focus on the servers first though. Check out the O'Reilly books for LDAP and "Linux in a Windows World" for guidance, but it really depends on how many people need to use those critical windows apps, and what apps they are. Let me know what type of apps you are talking about, as there may be replacements or documented WINE support for them. AD to LDAP isn't likely to be much trouble with only a few users, and any mail, file, and print services should be relatively simple to implement, whether you convert or use winbind to maintain AD integration. Having been on both sides, though, I would definitely prefer switching to LDAP first, as AD can give you plenty of headaches down the road. Also, regardless of which path you take, be sure to make use of NTP to maintain your clocks, since a small drift will eventually wreak havoc on anything using kerberos, and it might not be the first thing that comes to mind when something suddenly stops working.

Re:Look at costs, Servers first

[secolactico]

Having said that, you might be able to run your critical apps under WINE, and use Linux X clients to run it via SSH.

... after making sure with your software/support provided that this is a supported configuration. Otherwise, they might use it as an excuse when something breaks (even if it's not a wine issue) to wiggle out of fixing a particularly difficult problem (if they are anything like the provider of a company I used to work with, they probably sold you the Windows licenses and might not be tickled pink to see the OS upgrade revenue going away).

Don't rush things. Break in the users nice and calm. Set up sample workstations for each environment and ask them to give them a try and get their feedback. That way you'll be prepared to deal with the little (yet annoying) issues or even better, you'll be able to avoid them. For example, in Windows, the U.S. International keyboard layout differs slightly from the Linux version in the way they handle the entering of special characters. It's no big deal, but for a fast touch typist, it can really wreck your pace while you retrain your finger memory.

Good luck. If you do succeed, please post your story and let us know.

Re:Look at costs, Servers first

[linuxrocks123]

> The question came up... why should we anyway? Every new computer comes with Windows and all the drivers preinstalled. We just have to spend 30 minutes uninstalling all the crap that comes with it.

You shouldn't be using the default Windows install anyway. You should be using an image. Default Windows installs sometimes have viruses and spyware (no, really, they do).

> And then how much of your infrastructure will you move to Linux? If you're not using an app that requires Windows you might in the future. The next version of that app might be dependent on .NET 2 which may not work under WINE, or newer versions of the computers you're using might not work with the Linux images you currently have on hand... more work to get things going..

Well, first, if it's on .Net you would be using Mono, not WINE, which already does support most of .NET 2. Second, WINE is really only suitable as a stopgap measure today; if you're being forced to use it you should be planning on changing to an alternative program anyway. Third, WINE will almost certainly support the new version of an app that is already working with it -- eventually. So if you insist on abusing WINE in the manner you describe, you might have to hold off on upgrading for a few months but that's it. As far as the images issue ... you make new images for new computers, no matter what the OS is. (See above about default Windows installs.)

> You'll simply be less flexible. Not to mention if you get run over by a bus the next admin may not:
>(a) know what the heck is going on in that setup
If you document it properly, a competent admin will be able to understand it.
>(b) know Linux in general
If this firm would hire someone with no Linux experience to maintain a 100% Linux network, they have some serious HR policy issues.
>(c) agree with the setup
Why do you perceive this to be a problem?

The rest of your comment is ignorant drivel that doesn't deserve a response.

Why?

[NineNine]

As always, there's not enough information. Why do you want to do this? What are you trying to accomplish? What apps? How critical are they? If you want to switch just for the sake of switching, then really, you should be fired.

Re:Why?

[NineNine]

A vendor like Microsoft has no reason to be nice to a small- to medium-sized company, and this leaves anyone locked in to a Microsoft system vunerable.

OK, let's think about this realistically. MS is the largest software company on the planet, with a financial statement that rivals the largest companies on the planet. They're not going away any time soon, and their OS is used everywhere. There are tons and tons of applications of all kinds that will work with Windows.

Case in point: basic small-business accounting software. There are tons to pick from that run on Windows. You can go down to your local office supply store, and pick one of a dozen, and they'll all do the job. If you switch to OSS, you have about one choice: Gnucash, and it's mediocre at best (let's forget that it doesn't have some critical functionality, such as payroll). If Gnucash, a piece of software being written by a handful of loosely-organized volunteers in their free time, getting paid nothing, happens to die for whatever reason (very possible), then you're quite literally, SOL, unless you're a big enough company that you can pay $150/hour for programmers to re-invent the features that exist in a $100 off-the-shelf package. If you're using Quickbooks, and for some bizarre reason, Intuit shuts down (very unlikely), then you pick up Peachtree, or any of the others, export and import your data, and you're back in business.

I won't consider going to OSS because the inadvertent lock-in from having a lack of choice is very real. If I were to switch my company to OSS alternatives, there's no doubt about it, I would be "locked-in" to using what few options there are. "Lock-in" on the MS platform is unlikely. Sure, it could theoretically happen, but it makes as much sense to worry about that as it does to worry about a comet hitting the Earth tomorrow.

I feel that is is much safer for a smaller company without deep pockets to stay with mainstream software, as much as possible. Buy whatever is generic and does the job, then move on to getting to the part of the business that pays the bills.

Re:Why?

[ClosedSource]

"Bollocks. Avoiding vendor lock-in is a sound business philosophy for a cost sensitive small business. "

What is the dollar value of "avoiding vendor lock-in"? That's right, you did say it's a philosphy, so perhaps it's unconnected to the bottom line.

Re:Why?

[NineNine]

Oh, I own a tiny company, and I spend more than that. I've got about 10 machines, all with licensed W2K or XP. The kicker is that my point-of-sale software costs $1600/workstation, and I have 5 workstations, and support for this software is no more than two years. That's a *lot* of money for a six person company. But really, I have no alternative. It's a cost of doing business. There's no free alternatives to my POS software, and the OSS ones simply don't do what we need them to do (integrated credit card processing, integration with Quickbooks, Win 32 API to hook into our web site, etc.). So, I have to look at my business. My options are to spend $8K every few years of software, or try to run a retail store with more than 10,000 items and over $1M/year in sales with some kludged together OSS stuff that would take a *lot* more effort, and may not even be possible without spending about 20 years worth of licensing costs to pay somebody to develop something.

If I owned a white-collar business that used computers for basic word processing and email, then sure, it doesn't really matter what you use. But how often is that the case, in this day and age? My friend, an attorney (basic office job, right?), needed some good way to handle scheduling, contacts, email, etc. Of course, he went with Exchange. Why? After spending about 6 months looking for OSS solutions (and don't forget, he could have been using those hours to bill clients at $150/hour), he had lost a ton of money, he pissed off the other lawyers in the office with all of the software mess, and he looked very unprofessional when whatever he was using wasn't working, and he couldn't respond to his clients. Finally I told him to spend a hundred bucks a month on hosted Exchange service, and get on with his law business. Everything is running pretty smoothly in that office now.

Maybe, MAYBE if I ran, hmm... maybe a... hmmm... catering company, then OSS would work. All you need is some basic financial tracking (ooops... still no payroll), and something to print pretty estimates and invoices. But really, I can't think of a lot of businesses in this day and age that would be willing to do something so dramatic to save such a small amount of money (I spend about 30 times more on rent than I do on software).

Re:Why?

[NineNine]

A vendor like Microsoft has no reason to be nice to a small- to medium-sized company, and this leaves anyone locked in to a Microsoft system vunerable.

I'm about to spend a lot more money with MS, as we migrate our point-of-sale systems to MS RMS. They have very helpful salespeople that are willing to hold my hand even though the total bill won't be in the 5 digits, and they even are financing it for me. MS is actually very easy for my small company to deal with.

and this leaves anyone locked in to a Microsoft system vunerable.

Vulnerable to what? Give me a real world scenario. I just don't see it.

Re:Why?

[pogson]

"If you want to switch just for the sake of switching, then really, you should be fired."

No one wants to waste time and money switching for the sake of switching. Most open-minded IT folks understand:

• Windows is fragile compared to a UNIX/POSIX OS.
• Microsoft has a monopoly. It costs money to buy from a monopoly. Competition is almost always better. Let Windows compete on its merits by examining alternatives.
• There are tens of thousands of malwares out there looking for Windows systems.
• Microsoft likes to force huge costs when "end-of-support" for one of its releases is reached. This makes the locked-in folks believe Windows is relatively cheaper.
• Most organizations that switch to GNU/Linux cover their costs in reduced maintenance the first year. If specialty apps block the switch, perhaps the mistake made to go with those apps is better corrected sooner rather than later. Accept no app that is not designed with portability in mind. If a business is valuable, you do not want Microsoft or any Microsoft partners controlling it.
• What will you do when the hardware finally dies and your version of Windows cannot understand the new stuff you buy? You will have to replace everything eventually, anyway. Do it sooner rather than later and use FLOSS as much as possible to prevent a recurrence.

In my work, I helped a school switch when they moved to a new building. Previously they had on-site personnel to manage hardware and software. Now they can go many months without intervention. The conversion costs over and above the new hardware which they were going to buy anyway was $5000 and an hour long introduction to the new software. By now that cost is all recovered. They should not have to do major hardware upgrade for ten years and software is continually upgraded from the distro in a few minutes as desired. By not installing Windows, the cost of the IT system would have been cut in two except that they had a sum in the budget and spent it to get twice the capability. There has been no downtime since a faulty memory module was replaced after some weeks of operation. Earlier we did need to replace a driver for video. That was done in the off hours.

Granted we had no "specialty" apps, but we have way more software now than we did last year.

The librarian did insist on using proprietary software. The shrinkwrap was lost in the mail and her library is still not functional although we had a FLOSS web app available early on if she had chosen to use it. What is the cost of delayed implementation of a major component of our business?

Re:Why?

[TeknoHog]

What is the dollar value of "avoiding vendor lock-in"?

It depends on the timescale. Philosophies rarely pay off during this quarter-year, but they can make a big difference in the long-term survival of the company and the society.

Re:Why?

[RobertLTux]

What is the dollar value of "avoiding vendor lock-in"?

i don't know it depends on what %Nextversion of the app will cost you ie
Oh The new version has some many %shiny features that we have to charge 12X to upgrade but we will charge you only 10X if you upgrade NOW (and in ?months oh Old version is no longer supported and your upgrade window has closed so you will now have to pay 20X and purchase a legacy migration tool at $$$$ per seat)

Re:Why?

[greenguy]

WTF? A comment poo-pooing considerations of vendor lock-in gets modded to 4, and replies pointing out the importance controlling your own data get modded to 2 and 1, respectively?

Anyone who sneers as philosophy as being disconnected from real life (including "the bottom line") deserves to be modded into the ground. Exactly what do such people think philosophy is?

Re:Why?

[ClosedSource]

You can't just invoke the word philosophy and expect everyone to stop thinking. There's nothing unethical about getting locked in to a single vendor if it makes sense for your business. If your philosophy doesn't believe that the "bottom line" is key element in business, than your business will most probably fail.

It's OK to have a failing business and it's OK to have a philosophy that rejects basic principles of business, but philosophies get disconnected from real life when an individual's profession is in fundamental opposition to his philosophy.

Re:Why?

[sjames]

What is the dollar value of "avoiding vendor lock-in"? That's right, you did say it's a philosphy, so perhaps it's unconnected to the bottom line.

Avoiding vendor lock-in is a matter of risk management. There's an industry full of insurance underwriters that could probably put a dollar value on it, but since I'm not one, I don't really know.

However, It's not hard to understand the position vendor lock-in puts a business in. When a business has it's critical data locked up in a proprietary format and one and only one vendor can grant them the right to continue accessing it, that vendor is said to have the business "by the short hairs".

There are always costs and pain associated with switching vendors, even in the absense of lock-in. Compatability goes only so far. The question is how much is that pain worth. A vendor that practices lock-in tactics generally knows about how much and calculates the ongoing costs to fall just short of that mark FOR NOW. They are well aware that they can, in practice, extract many times that much by spacing things out far enough. All they have to do is present you with a bill for 80-90 percent of the cost of switching and make sure you percieve the next shakedown as non-existant or in the distant future (better yet, as coming from someone ELSES budget next time).

Of course, if you can afford to look just past the annual report, you'll save a substantial sum of money by accepting the pain NOW and reaping the benefits later.Simple economics will tell you that when multiple vendors offering approximatly the same value have to compete for your business, you win.

So, it's not always directly connected to the bottom line, particularly to the bottom line THIS quarter or THIS year. However, it IS connected indirectly and the effects are real. In part they are masked these days by the high prevalence of the lock-in. It can be hard to realize how badly you're doing when everyone you can see is in the same boat.

All of your issues are no problem.

[Shaman]

We have a single W2K3 system which serves up a couple of legacy apps over RDP (Rdesktop) and integration with Samba, etc. has gone well for us. The standard KDE applications work fine although you do have to choose your distribution, largely because Flash can hang and/or crash Konqueror on a regular basis (blame Flash, not Konq).

The only issue we have run in to is that Windows will only let you log in with RDP so many times before it will blacklist your machine's hostname for not having a genuine MS license. It's a pain but we just more or less randomize the hostname regularly. Good old Micro$oft... they won't even let you administratively remove the blacklisting without delving into the Registry (haven't tried that, but I figure it must be possible). This happens infrequently, by the way, W2k3 will probably accept a good 100 connections before it whines.

Re:All of your issues are no problem.

[richie2000]

He *is* legal, his client box runs Linux (see the rdesktop reference.) Yeah, but MS requires Terminal Services Licenses for the clients. These come with XP but would theoretically need to be purchased from MS if connecting with other clients. In addition, you probably need to have enough CALs too, depending on what the servers are being used for (for example, a Win2k3 SBS comes with only 5 CALs). I had to research this whole scam^H^H^H^Hscheme back in the Win2k Server days and it's a total bitch. Apparently it's even more convoluted in 2k3...

Re:All of your issues are no problem.

[Sparr0]

Fortunately most of us live in a country (the USA) where MS doesn't have the authority to "require licenses".

Start with your applications.

[jonadab]

> however a couple of our core applications run exclusively on Windows

Then that is where you have to start.

Yes, you could insert a couple of Linux systems in side roles that don't require them to run the core apps, e.g., a DNS server here and a CGI server there and so on and so forth -- and that's likely worth doing for its own sake -- but if you want to migrate entirely off of Windows, you've first got to migrate to all cross-platform applications.

Re:Start with your applications.

[wynler]

Absolutely 100% correct.  If you have a small company, and applications that require Windows.  You don't switch to Linux.

VMware Server, Converter Beta

[Semireg]

Next year much of the server and desktop hardware we run will need replacing. Migrate your servers to virtual machines. You can do this for free using Cent OS as the host, and VMware Server (free) software to run virtual machines. The VMware Converter (now in beta) will allow you to p2v, or migrate physical-to-virtual machines and this is done while the source server is powered on. So, regardless if you're going to Linux right now, you can make the jump to hardware-agnostic VMs with just a few clicks, and no extra money spent. Right away, you'll gain flexibility by utilizing your new hardware more efficiently. Good luck!

My Office

[Sparr0]

My current boss, a close friend of mine, single-handedly began a FOSS migration in our 3-location 100-desktop 20-offsite-laptop-user office about a year ago. I came on board about 3 months ago, almost through the first stages of the process. We now have 99% of our users on OpenOffice (one holdout, and I am going to fix his missing feature ASAP to get him off Excel), and 100% migrated away from IE+Outlook (most on Firefox+Thunderbird, a few people requested Mac desktops and are using Safari+Mail). We transitioned to Open Directory on an OS X Server with nary a hitch, with the added bonus that OD supports LDAP which means it plays nice with all of our new extranet and internet services (LDAP login to our helpdesk, CMS, etc).

Eventually Windows XP will lose support and we will have to consider sticking with unsupported XP, or moving to Vista/Fiji/Vienna, or a complete migration to Mac, or a final alternative that I am starting to push slowly up the list of possibilities... Linux. My boss is a Mac user, he dislikes many of the problems with Windows. He had the popular misconception that Linux is hard to install, hard to maintain, and hard to use in general. My first day, when provided free reign over my own desktop, I let him watch me go through a Kubuntu installation. Cleared up all that nonsense right quick. From a blank hard drive to a better-than-Explorer GUI, with both of our network printers completely configured, desktop shortcuts to our network shares, Firefox and Thunderbird installed as well as a GUI terminal (we have legacy apps requiring telnet to our SCO UNIX machine), all in under 30 minutes, and without touching a text console.

Running actual GUI Windows applications in Linux CAN be difficult, but often is not. There is a VERY good chance that they will 'Just Work' under WINE or Crossover Office. If you need terminal services functionality, rdesktop has worked great for me. There is also the VMWare/etc option, if the programs are old enough for the perfomance hit to not matter (and if you're developing "core" applications that only run on Windows TODAY, then youve got other problems).

Provide the complete analysis first

[tftp]

One computer costs $1,000 in hardware. One employee costs $120,000 per year, with burdening. One "mission-critical" application costs anywhere from $800 (AutoCAD 2007) to $5,000 (Inventor 11, non-pro.) One WinXP Pro license costs mere $150 even if you buy it at maximum cost, as a retail box. Now, aren't you putting the cart way ahead of the horse? A single wasted hour of any of your employees' time (or your own) will cost as much as an XP Pro license. Have your numbers straight before switching, and have very good reasons to switch.

The problem with businesses is that they are not very open to OS theology; businesses just want to do what they are doing, and if the job requires computers and OS and apps and stuff, well - that's just the cost of doing business. It will cost money to run a Linux shop, and it will be probably *more expensive* to run a Linux shop, considering that every Windows app -- that normally would be "install and run" on any Windows box -- becomes a WINE nightmare, to see where it crashes and how to work around those crashes. Do you really want to buy a $20,000 app (there are plenty of specialty apps in this price range, all mission-critical) just to find out that no, it won't run under WINE, and no, vendor support in such environment is not provided. Do you want to lose the support on such an expensive app? You are risking not just your job, you are risking jobs of your coworkers too - if the company loses a contract because of OS troubles then some employees may need to be laid off, starting with you, of course.

If you have dreams about using RDP for those few apps that you must have on Windows, it depends on what those apps are. Some apps do not permit running under RDP because that would be inviting to buy one copy of an app and then have the whole company to access the server and run the thing. I personally know of some examples, so check before you buy into it. And other posters already said that the cost of a terminal license is as high as WinXP, and you have all the eggs in one basket (server.) Server dies - the whole company stops; are you OK with that?

Again, businesses don't want anything that deviates from tried, tested and true path. Cost is not a concern here; labor and apps cost uncountably more than the OS. If you want to migrate, you still can do that; I tried myself, starting with a 3-man company, and guess what eventually happened? Once we started growing, the total cost of maintenance of a mixed network shot through the roof (and disappeared among the stars.) Now we stick to Linux on firewalls, and Windows XP everywhere else. We do use Linux on our embedded systems, and it's perfect there. Desktops are a different matter.

Re:Provide the complete analysis first

[Anonymous Brave Guy]

The problem with businesses is that they are not very open to OS theology; businesses just want to do what they are doing

Some of us would argue that this is not a problem, but a feature.

Look at the "why" first.

[mrscott]

Ok - since this is Slashdot, I expect to be thrashed for looking at this from the business perspective (I'm a CIO with 13 years of IT experience). The first question to ask yourself is this: "Why?"

Ok, I'll be the first to admit that there is a tremedous lure to FOSS software and have rolled it out myself in a number of situations, but not to desktops. I've replaced web servers, database servers and Windows file servers with servers running Apache, PostgreSQL and Samba. However, before I considered something like this in my current environment, I'd need to do a serious cost analysis that went way beyond licensing costs. For example, what will this mean to the user that has been using Windows and MS Office for 10 years? And, you mentioned that some of your core applications are Windows-only affairs. Sure, you can use RDP/Citrix to run these apps, but then you're throwing the Windows licensing costs into the mix. Not to mention the possibility that your apps won't like running in this way.

So, how much is your infrastructure *really* costing you?

How much would retraining cost?

How much would it cost to possibly have to give up your core vendor support due to running in an potentially unsupported configuration?

This may sound like I'm anti-FOSS. Actually, I'm not - I love FOSS in the right situation. WHat I AM against is FOSS for the sake of FOSS. While I "grew up" on the IT side of the house, I'm a big believer in the business needs dictating IT's role and responsiblity rather than the other way around.

My advice: Think this through before you put a lot of time into it. You may end up saving a whole lot more (not just money) by sticking with what works.

Re:Real world vs. fanboy fantasies

[altstadt]

The plural of anecdote is not data -- Frank Kotsonis

<anecdote>

Now, here are the facts as they're found in ONE PREVIOUS PLACE OF WORK:

We had roughly 150 people working in a branch office, 110 of which were a mix of hardware and software engineers. The rest were either support or upper management.

We had roughly twice as many computers as people, with the computers in the lab area shared among many people depending on who was using a bench on any particular day.

About 80% of the computers were running a couple of Unix variants, mostly Solaris. The rest of the computers were running Windows.

We had 3 full time IT people who had to support all the workstations, servers, and communications equipment.

• The IT people reported that 80% of their support tickets were for the 20% Windows machines.
• Since we didn't have root access to the Unix machines, many of the remaining 20% support tickets were spent in either shutting down Unix machines so we could move them to another bench, or for installing new hardware and/or software.

</anecdote>

I have yet to talk to anybody who has actually experienced a situation where Windows support and development costs were less than Unix (or Linux) support and development costs for the same staff at the same location. I figure these places must exist, because SEVERAL INDEPENDENT RESEARCH INSTITUTES seem to stumble over them all the time. I'm glad I've never worked at any of them though.

Re:Real world vs. fanboy fantasies

[nacturation]

You didn't explain why you would want to migrate your shop to Linux in the first place. You even mention that the software you need is Windows-only stuff, but you want to make things complicated, difficult and expensive by running this Windows software on Linux virtual terminals instead of natively! Those first two sentences contained some great advice. It's too bad your post turned into what looked like a bunch of [independently researched] BS numbers.

However, if I were to add to that first bit as a reply to the submitter, I'd seriously consider the question of whether or not this small shop can continue on servicing a Linux deployment with a complex mix of Windows/Linux after you leave. After all, you don't plan to work there forever and given that you have to ask others for advice, how likely is it that:

A) you can seamlessly make the transition yourself; and
B) someone else can easily pick up where you left off?

Unix-based servers are absolutely great and typically rock solid at doing server kind of stuff... much more so than Windows presently is. However, I'd actually advise you to stay with Windows. It's what a lot of people know, you know it currently works, and unless there is a serious compelling reason why you can't just continue with the status quo, it's the cheaper to use what you have than try and make changes with potentially unknown complications.

If anything, I'd setup a parallel network running Linux and host some services off of that, gradually migrating services one at a time over to it while you transition off. And if things go south and you run into issues you can't resolve, you could always pull the plug and you still have your original Windows network.

Re:AD

[Realistic_Dragon]

Oh, and when putting machines out for testing you have a good opportunity to help manipulate the users.

Make the test machines pretty spiffy. Get some flat panel displays for example, if you haven't already got them deployed. Draw lots for who gets the 'first upgrades' rather than allocating it out like it's work.

Properly set up (if your office is anything like mine just set the default screensaver to the 3D matrix one and make them dual screen machines) you will get huge enthusiasm for 'the upgrade' rather than bitching about how everything is now insignificantly different.

Re:Have your numbers straight

[anomaly]

Ok. You're right. That means that 2 hours of employee time make up for the cost of one Windows XP license.

Please don't misunderstand - F/OSS provides LOTS of great software, but I don't see any way you can pencil the cost of Linux as a desktop replacement for Windows. Linux makes just about everything possible. (FWIW, I have been a daily Linux user since 1994.) Just because it's possible doesn't make it a good idea. Just because it's cool doesn't make it make any business sense, either.

All of the software/hardware vendors work their butts off to make sure that Windows compatibility is met. This doesn't mean that they do it well, but they don't care about any other OS. You can care, and if you select peripherals well, it won't be any issue at all. What happens when one of your "important" users goes out and finds a great deal on a digital camera/printer/trackpad/some other device which is completely unsupported in Linux?

It's not worth fighting the battle for the desktop. Linux is not complete enough yet for non-technical users to have. Linux on the server makes great sense, and I highly recommend it. (Although at home I just migrated all of my services to OS X.)

Respectfully,
Anomaly

Moving small organizations from Windows

[westlake]

I am looking for creative ways to introduce Linux as my desktop and server OS of choice

Hold up, there, cowboy. That is the wrong question to ask.

The systems and servers aren't your personal plaything. They are there to meet the needs of your employer. The small organization. The all-Windows shop.

There are often reasons for choosing the proprietary app. The predominant OS for a business of your size or type or location. Reasons that are not always narrowly technical, not always narrowly economic.

Re:Warped Analysis

[tftp]

Software costs are a burden. Employees are productive assets.

Software and hardware costs, rent, business licenses, salaries and taxes are your business expenses. It does not matter what names you use; it only matters what you pay for. If you rent a tool, it's out of your pocket. If you hire an employee, it's out of your pocket. Money-wise they are the same.

The rest of your analysis is based on the presumption that Windows works. If that was true, no one would be considering a migration.

Modern Windows works, that's not the problem. IMO, one of primary motives to migrate to F/OSS is costs of licensing. Windows-only infrastructure may be expensive if you go beyond the desktop and start buying PDC, BDC, TS, SBS, Exchange and other servers that MS will happily sell you. That's where the real cost is - server license, CAL licenses, TS licenses... start counting. That's what I consider a motivation. Cost of a desktop OS is nothing. Cost of several Win2k3 servers, each with full complement of CALs for all your employees, can be devastating.

If your core apps run on Windows...

[slk]

Then you probably need to be running Windows, at least on the client.

I have a day job as the head system administrator for a medium sized but very high-tech non-profit. We run Macintosh (OSX) clients and Linux servers because they do what we need to do, and do it well. I have also been working with Linux and various other forms of Unix since 1994 (this includes using Linux and/or FreeBSD as a primary desktop OS since 1994. LaTeX works fine as a word processor if you know what you're doing.)

I also do consulting work for several smallish companies, and they all run Windows. It's really simple - if you need good 2D CAD software, you need Windows. If you need a modern multi-user accounting package that can do strange things like payroll and integrate with direct deposit, you need Windows. If you need a *good* spreadsheet (no, OOo calc doesn't count), you need Windows or OSX. If you want to run all of this on one desktop operating system, you need Windows. Crossover Office, WINE, VMWare, etc. aren't going to convert many small businesses; they want less complexity, not more. (some of these clients have Linux servers - network edge, multiprotocol file and print services, web apps, etc. - but they are close to 100% Windows on the desktop)

I think that you could convert a LOT of small businesses over if you could get a Peachtree or Quickbooks port for Linux. However, for small business, you don't stand a chance until you get *good* accounting software. OOo calc not sucking would really help too; lots of businesses make very heavy use of spreadsheets. (OOo Writer sucks, but so does Word. OOo Impress is adequate, as it's all pretty much PowerPointless anyway.)

If you're looking for long-term savings, I'd suggest considering Windows TS clients (use your old XP machines/licenses/etc), and a Windows 2k3 server terminal server. It won't be all that cheap to setup initially, but you will be able to significantly reduce your maintenance headaches.

Look at the business needs, and pick technologies that meet the business needs. Make technology work FOR your business; I've see what happens when you flip that around, and it isn't pretty.

Almost total Linux shop with 1000+ employees

[SQLz]

I actually don't work for the group in charge of maintaining the systems, but I know a few things about how they are maintained. Basically, all systems have the same exact same RH4 image and sync up against an internal yum repository for software updates. There is basically zero maintenance for each machine besides that. Users can't write to the hard drive, all data is stored on netapp filers. When you are hired, you get really basic classes on how to use KDE, the internal wiki, Open Office, get on mailing lists, etc. A caveman could pass these classes.

We have over 7000 linux machines and 4 people to maintain them, plus 1000+ technical and non technical employees. Using Linux saves us millions of dollars, which pays for a couple of those netapps. The thing is, Linux just works, not to mention the vast amount of free software that is available for it.

Truthfully, and its a sad truth for some people, anyone who says Linux isn't ready for the corporate world has no idea what they are talking about. Its been there for while.

Re:Real world vs. fanboy fantasies

[altstadt]

Good point, however your assumption is incorrect. Most of the Windows machines were co-located with a Unix machine on the same desk, with a few more running dedicated Windows-only applications in the lab. Roughly 95-98% of the staff were regular Unix users. I can only think of three people who had only Windows machines at their desks. One admin assistant had started in the department using only a Unix machine for a few years (typing reports using *roff).

As I stated in the original comment, we didn't have root access, therefore we couldn't solve our own Unix issues. The IT people worked on 100% of the Unix issues which came up. Many of these issues were nothing more than getting a machine cleanly shut down so it could be moved.