Let's kick off this forum: open source in the workplace

I personally do not have end-user experience in a large environment with linux. Servers are a different story. I do know of a few companies and educational systems that have gone to end user systems though.

The main pains were as in my previous post, obscurity. I think Gibson music is exclusively Linux now and a west coast clothing chain and Munich. Japan and China seem to be making a big investment as well. Obviously these are just examples!

Do I think it is desktop ready? Yes. I think RH and SuSE (UnitedLinux) have got to a point where their systems are very easy to use. I was very impressed by Yast when I tested 9.0 and 9.1. SJD looks very promising and I can't wait for Looking Glass to be released (presumably SJDII ??)

The thing I hate about SuSE, Drake and RH is lack of configurability (or the hurdles associated) with their distros. Kernel configuration is a bear on RH and SuSE if you are not using their sources. As a server and embedded type person I like purity like Slack, Gentoo and *BSD.

BTW, sorry about the last post, I wasn't jumping on you.

Yep, Chort, I agree with you when u say that Linux is Enterprise ready as a server platform not as desktop.... Mainly not only because of a "lack of direction" for the one billions linux distribution out there...
I think that, for all medium-large companies, the "change of the desktop" means that all the people that worked for a long time in a MSwindows environment should learn to use a different one. Now, not the skilled technical or whatever else similar, the problems are all that people that make a "day by day" office work.....
If u gonna want to change u need a strong, simple, fast new desktop OS
...yep, I know, maybe one or two distros may do the joke.....

PS when I say that *ix isn't desktop ready I mean in an enterprise environment....Now I know well that if u want to use Linux at home it's fantastic even if you aren't a guru....
I work in a large fashion company and in office I use both MSW and slack....To go throught ms desktop and linux is not so quickly .....

First, thanks to Jeremy for this forum. I've been looking for one that focused on Linux in the business world.

I've been on the help desk front lines for over ten years now, and in corporate IT for nearly twenty. I'm as new as I can be with Linux, having done only one RH9 install on an aged Compaq Proliant 1500 (as a desktop, not as a server). And installing it is all I've done; I haven't had time to do anything else with it yet.

I have to agree with chort about the "choice" issue. As I'm trying to learn this OS, I frequently get bewildered by the number of choices available. And not just the choices for apps but for things Windows administrators take for granted, like text editors and file managers. I've quit asking for recommendations of various applications, since the answer is frequently "Try one; if you don't like it, keep trying others". This is a good answer if I'm doing this for a hobby, but isn't practical from a corporate support approach. Corporate IT departments view this as a waste of support time.

When I evaluate new software for Windows systems, I use publications like PC Magazine and PC World to help me select five candidates or less for further evaluation. The presence of these products in publications of this stature tells me these products are from major developers with some form of product support available during Eastern Standard Time, even if I have to pay for it. I then go to manufacturers' web site to compare the features and narrow the field down to two or three for testing. I know if the application says it will run with specific versions of Windows, it will run with those, and no others. I know I'll be able to install it by double-clicking a Setup icon, ignore the license agreement, and answer a few simple questions. Everything I need to run it is included; I don't have to go hunting for additonal files or programs the developer assumed I had. He know what I've got because he tested using Windows.

With all the various Linux distro's, is it practical for a developer to test with all possible combinations of kernel builds, productivity suites, desktops, browsers, file managers, etc? There may be comparable publications for Linux apps in corporate environments, but I still green and haven't located them yet. And I have never posted a query in a support forum without at least one response telling me to switch to another distro or application package. That isn't support, that's a debate. My VP doesn't want to hear that my support system for the corporate productivity suite is posting a query in a forum and waiting for a response.

One of our prime criteria for software selection is if the software is being used by our customers, suppliers, and other companies in our industry. Compatability with the supply chain is critical in corporate environments. In some cases customers mandate in contracts what software we are to use when we provide products to them. This is especially true when the customer is a government agency.

As someone else pointed out, I've found Linux documentation to be less than satisfactory. I shouldn't have to learn an navigation system just to read an INFO file. I suspect the documentation is so unfriendly because writing help files, documentation, etc. is a boring chore that no programmer wants to do for free when he could be developing slick code. That's why commercial apps pay technical writers to do it.

Don't let anyone tell you if you load Linux, a GUI, OpenOffice, and Mozilla that your end users will not be able to tell the difference. No one in IT ever got fired for buying IBM or underestimating the capabilities of the average end user. Most end users know how to use their computer by memorizing keystrokes and mouse-clicks. Swap two icons and they'll be paniced when Word starts instead of Excel; their computer is "broken". I've seen end users panic because someone else used their computer and changed the wallpaper. They tell their manager the computer is broken, and he wants to know when the support staff is going to fix it.

End users aren't interested in the wonderful capabilities of Linux (or Windows), or the choices available through the world of open source software. To end users and corporate management, a computer is just a tool, like a hammer or a car. Management wants to know that employees who learned how to use a hammer or drive a car somewhere else will be able to apply those skills with a minimum of retraining, preferably none.

<<He steps off his soapbox, ready to dodge Spring's first overripe tomatoes.>>

That post is all fine and well with the exception that all the choices are completely irrelevant. Sun, SuSE and RH make specific versions targeting end-user desktops. I've used almost all of them for testing purposes. If you've used RH9 then you should have noticed this. RH9 comes with a specific set of solutions for the end user. Everything is tightly integrated on the desktop.

Linux has plenty of magazines and online documentation to help you with these choices as well.

I don't understand why someone would say, "gees, theres 50 distros to choose from." Who cares. We all know there are three or four viable candidates. There are only a few candidates for productivity software as well. Take Ximian for instance.

Ximian even has a management system for wide scale deployment and package management.

For crying out loud, what do most end-users do anyway? How about, check mail, view documentation, powerpoints and use web based applications and maybe a special application. Most corporate environments use mainframe or centralized apps over a connectivity client for their main work anyway.

The fact of the matter is this, Linux doesn't match Microsoft 100% for wide scale deployment, but departments better realise that they should go throught the slight pains now of switching to alternative environments. If they do this then the support will rapidly catch up with money invested. If they don't do this, then when Longhorn goes wide scale Microsoft will virtually own the Internet and all corporations pocketbooks when it comes to liscensing.

As I said, I haven't used RH9; I've just barely been able to loaded it. Which desktop has everything integrated, the Gnome or the K? I'm very new to this; I didn't know that there are only three or four distros suitable for corporate use. Which ones are they? As to productivity software, I hadn't even heard of Ximian until today. But again, I'm new around here.

Please recommend a magazine for the Linux newbie, preferably one for somebody with an Wintel or professional background. As to web sites to help with software choices, I'm wide open to suggestions there too.

This is a country uses one primary source of automotive power, relies on a small pool of media conglomerates for info and entertainment, and picks it's governmental leaders from only two parties. What makes you think it will care if it's computer operating system and apps only comes from one company? Notice I did not ask why they SHOULD care (and they should), but why do you think they WILL care? Corporations are slow to react; the bigger they are, the slower they are. Many of them operate on the principal of "if it isn't broke, why spend money to fix it?" They won't spend money in hopes the support structure will grow; corporations don't work on the "Field of Dreams" business model. Emulating Windows on the desktop isn't the key to displacing it anyway; a killer app is needed, comparable to what Lotus 1-2-3 did for DOS, or what web pages did for the Internet.

As for Longhorn, Microsoft is not seeing XP deployed in the numbers it hoped. While it shows up on home machines, corporations are not buying XP licenses for existing machines. There just aren't enough differences to make it worth the trouble. When they buy new machines, they are often installing 2000, despite having just paid for a new XP license. Many networks are still running NT 4 servers, despite the presence of server versions of 2000 and the new 2003. MS Office sales are also below what MS had hoped; corporations don't need the new features (or the expense of upgrading operating systems to support it), so they are staying with Office 2000 or even earlier. Yes, MS discontinues support for products after three years, more or less. But many companies are taking the approach that any problems will have already been solved, and don't care if it isn't supported. Corporate IT buying policies regarding Longhorn may be the same as their attitude toward other recent MS introductions.

MS already effectively owns the Internet. If web based applications are the growth market, all the developers seem to think it's important that the app be IE compatible. How often do you see a web site that says "Best viewed with Mosaic"?

Opinions are definitely important and everyone seems to have one. Mine is that Linux isn't Windows and hopefully never will be. If someone wants it to BE Windows, then why are you switching? Also Linux is not easy for the beginner and is not intended to be. There are plenty of tools to make it easier and its all up to the administrator or end user. I love Webmin!! It's makes administration easier, but far from easy! You still have to learn it. If you have time to research Windows programs, you have more than enough time to investigate a text editor that suits you (I like jed which probably won't be in most Linux users top 5).

I can tell you that when the NT wave came in, old Unix admins struggled just as much initially adjusting to the switch. Use Linux,...learn Linux,... enjoy Linux! Nobody is going to hand you a magic pill to make you a Linux guru and we can't make you like one program over another. Learn more about Linux and you will see that choice is a good thing.

I don't want it to be Windows. While I would prefer a GUI, I can learn to work with whatever is necessary to get the job done ON SERVERS; that's what I get paid for. The number of choices available is causing me some sensory overload, but I don't yet have the resources yet to filter them out. I'm hoping for some guidance in this forum; maybe a link to something like "Linux for Windows Admins and Other Dummies".

If Linux is "not easy for the beginner and not intended to be", will it ever grow beyond the server room?

My point is Linux and open source apps need to look and act like Windows, Office, Photoshop, Acrobat, AutoCAD, and other defacto corporate standard software if they are going ON THE DESKTOP. Otherwise end users are going to need retraining and will lose some productivity while climbing the learning curve. And unless the corporate world standardizes on one or two programs for each application, the skills learned at one company may not transfer to another.

There's much I don't understand about the open source model in general and Linux in particular and how they impact corporate IT. I'm still confused over the "free as in speech, not as in beer" phrase; what -exactly- do we mean by "free"? Is there a synonym that will explain the concept better than the word "free"? I'm used to thinking of an OS as a coherent whole with a reasonably consistent interface, not having basic things like print sharing, file management, resource browsing, and user management made available as separate modules from a variety of third parties with different interface approaches. I've only seen one distro; is there an easy way to tell what is included in each without having to load them all? That would certainly make it easier to determine what packages / modules are installed the most frequently and therefore have the best chance of becoming standards.

Why do so many people suggest I like / enjoy / love this OS? I don't like (or dislike) Windows, DOS, or VMS. An OS is like a workbench; its got to be present, but the screwdrivers and drills (applications) are what you work with to produce something.

The reference to learn and love Linux has nothing to do with not liking MS. I've been playing with Linux for over 6 years now and still am far from a guru. I stated that only because the more I learn about it, the more "WOWED" (if that is a word) I get. I like,...no,...LOVE the fact that all these tools are at my disposal and don't need the approval of anyone to get a budget to test them. If I get the preverbial "wild-hair" I am free to act on it (for testing of course).

Please don't think I'm trying to be the heavy hammer here. I do wish everyone could grasp the concept of Unix from the start but that is not the case. I do have an issue with newbies always wanting to make things more "newbie friendly". That usually ends up leading to tons of assumptions by the developers and ends up making more work undoing everything for the none-newbies. I am a systems admin for a small engineering firm. I've been in your position, but being exposed to the ins and outs of Linux over the period of a few years, I wouldn't change a thing. Sure, there is a lot to understand and comprehend and a passing interest in Linux isn't going to be enough to administer it. Only exposure will help. As far as documentation, it could be better, but it out there! Let Google be your guide!

On the flip side you stated that making things easier for the newbie will keep Linux in the server room. I agree 97% which is close enough. With my current knowledge, Linux on the desktop would be an administrative nightmare for me but that's my own problem. Quite a bit would need to change, but as long as AutoCAD is Windows only, nothing would matter at this point. We are locked in on the desktop.

Lastly "free as in beer" means you can give Linux or programs away to whoever you want free of charge and not get in trouble. "Free as in speech" means you are free to change the code to suit your personal needs an not have to pay anyone a dime for the code they wrote before you changed it.

As Martha would say,..."and that's a good thing!"

Also Linux is not easy for the beginner and is not intended to be.

Which is exactly why it's not widely deployed on corporate desktops. Enterprises don't want to incurr the massive cost in money and lost productivity of retraining users, and further they don't expect users to know anything about computers beyond moving the mouse and clicking the buttons. You forget that people who are accountents, clothing designers, marketing associates, etc aren't really interested in computers and they only want to know enough about the computer to do their job. In fact, they would prefer that computers are so intuitive that they can just read the prompts and not need to have any knowledge.

You can't have it both ways. Linux can't both be "the ultimate hobbiest OS" and also "the major desktop OS" too. Face it, office workers are not going to suddenly want to learn lots of things about computers, nor have the time to do it. The only way companies are selling Linux is by making it more simple and less cluttered (see SJD). No company will ever deploy Debian or Gentoo as their desktop. Apparently you didn't read the rest of the thread.

ARRRGGHH!

Yes I agree with you on the expense to retrain, but I'm referring to newbies using Linux! Take a person that has never used computer in their life and stick them in front of Windows and Linux. Almost unanimously, they will say that computers should be easier to use regardless of which machine they are in front of. Someone that is willing take the time to learn something beyond "push the big button on the front" will see that its not that hard.

This is exactly MY point. Most newbies mean by "make it easier" that it should work just like or similar to Windows only more reliable! If you take a second to unlearn Windows its no harder than anything else.

I will repeat and say that Linux is not for someone that knows nothing about how it works. To me it's similar to telling my grandmother to go build the Sears Tower. Windows knowledge is not transferable to Linux for the most part although every newbie with MS experience thinks it does. It's only because they haven't been exposed to something different. I don't see choice as a hindrance to its acceptance I agree with the previous poster that the lack of commercial applications is the problem! Sit me in front of AutoCAD for Windows and AutCAD for Linux (when its available, if ever) and regardless of the OS, I'll be able to use it! See what I mean?

Not only have the read this entire thread, but I've been participating since the beginning. Again, I'm not trying the be the heavy hammer, but my opinions I guess are different than others in this thread.

Take a person that has never used computer in their life and stick them in front of Windows and Linux.

Exactly. It's called the law of primacy. Most people know how to use Windows to do what they do. That is the largest hurdle with switching to something else.

For instance, my wife does not use computers at all. She doesn't even know how to use windows very well. But for the longest time at the house she had to use my BSD box to do everything, and she learned it just fine.

She used Staroffice to write letters and other documents and uses Firefox to surf the web. She even likes to play the games.

IMO Linux is not that hard to learn. As a matter of fact, when I show people how to use it most of them are very intrigued. I've had a very high success factor.

I've seen Windows servers in corporate use get screwed up and it took forever to get them fixed or you had to recover from tape. Unix boxes are much easier to fix. Everything is a file. In Windows you have the mystical registry and if something gets borked you're reloading.

It's unrealistic at this point to assume any significant portion of future users are starting from scratch. There are definitely new computer users coming in, but they're coming in as kids and growing up. The place you want to train them is in grade school, not once they've already reached the work place.

It's a moot point now whether it's more difficult to learn Windows or Linux if you don't have any existing computer knowledge, because face it most people (in the workforce) these days do have Windows knowledge.

Now, the real issue isn't making it "like Windows" for the sake of emulating Windows, it just happens as a natural course because the most lazy way to make your GUI "intuitive" is to copy what the user already knows. It can be argued (and Apple does so, vigorously) that if you would only focus on making the interface really intuitive, it won't matter if it's different--so long as it's very easy to learn and use, it's not a problem. The works well for Apple, but of course it was Apple that Microsoft ripped off in the first place when designing their current evolution of GUIs.

For the time being, "intuitive" will continue to be associated with "like Windows" until someone really gets their act together and makes a Linux GUI so intuitive and easy to manage all the essential components that no one cares if it looks different than what they're used to. Unfortunately, this will have to be a really outstanding GUI because humans always hate change and hate the unknown, so they will even complain about changes that are for the better.

For servers it's quite a bit different, because there has always been a much higher diversity in the platforms being used, so most server admins know at least a little about an OS other than their primary (except for a large group of very poorly educated Windows admins--thanks for nothing you lousy "tech schools"). Lots of environments use Windows + Solaris/AIX/HP-UX/whatever, plus a lot of network appliances with their own proprietary interface.

By the way, commercial support will continue to be one of the biggest hurdles to cross (I think everyone recognizes this) because it's the old chicken vs. egg problem (with some added difficulties). First, the hardware manufacturers won't release Linux drivers and the software companies won't port their software to Linux, until there's a very large following that makes it worth their effort (in sales) to do so. The wide adoption won't happen until the driver and application support is as good as Windows. Hmmm.

To add to that problem, GNU/Linux gang (well, most Open Source projects, actually) make it really difficult to hardware vendors to contribute anything besides source for drivers (with binary modules being heavly frowned on and actively discouraged). This makes the hardware vendors even less willing to cooperate. With the proliferation of proprietary designs, such as the nForce hardware from nVidia, the 802.11g wireless chipsets, and the Gigabit Ethernet chipsets, it's really essential for hardware manufacturers to share their documentation, or at least source code for drivers.

Please see the title of the thread; we're not speaking of total complete computer virgins. We're discussing the feasablility of migrating corporate users to Linux and open source apps, when they are experienced with using Windows, Office, and other Windows-based programs that are accepted as corporate standard applications. While these users may not understand the technology behind the Windows software they use, they are comfortable with it, productive, and uncomfortable with change.

I do agree with your point about the lack of commerical applications. This relates to another question I have about the open source model. I don't understand why someone would take the time to write an application without compensation when they could be getting paid for doing something they apparently enjoy. Perhaps I'm just too mercenary. It makes me wonder if their skills aren't good enough to get paid as professional programs. This in turn makes me wonder if their code is any good. I am NOT saying open source programmers aren't any good; I'm just trying to figure out their motivation and the business model.

My last posting was in response to ghight and jcookeman, not chort. I read the last post on the first page and inexcusably failed to read the new second page before I replied. My humble apologies to all concerned for my inability to follow a thread. I will do penance by installing WordStar on a Atari II.

So, according to you guys there is absolutely no way anything other than Windows will be in the corporate world, although there are very intuitive gui desktops from RH, SuSE, Mandrake and Sun, until they are exactly like Windows? What's the point?

Again, we all know that people don't like change that much, but it is something they will groan about for a few days and everything will be back to normal. We're talking about a desktop os!