The impact of social media on the wine blogosphere

Wine Blog on Ulitzer

Subscribe to Wine Blog on Ulitzer: eMailAlertsEmail Alerts newslettersWeekly Newsletters
Get Wine Blog on Ulitzer: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn


Wine Authors: Mehdi Daoudi, OnPage Blog, Jyoti Bansal, Scott Allen, Ben Bradley

Related Topics: Apache Web Server Journal, Wine Blog on Ulitzer

Apache Web Server: Article

Linux on the Desktop: Bringing Linux into the Corporate Environment

Linux on the Desktop: Bringing Linux into the Corporate Environment

Linux is coming to a desktop near you. The question being asked has gone from "Can it happen?" to "When and how do we get there?" This article is a high-level overview of some of the technical issues revolving around your corporate Linux desktop.

Fear, Uncertainty, and Doubt: The Current FUD

Before any corporate-sponsored Linux-based project gets under way, the current crop of FUD should probably be addressed. It might save some time later, and the mere fact that it comes up is a testament to the success of those who are spreading it. There is so much of it that there's no way this article can be remotely comprehensive on the topic. You'll need to evaluate what has come your way: ask the principals what they have heard and address it so that everyone understands that this is not another backroom project. For the purposes of this article we'll largely assume that we're talking about bringing a Linux desktop into the corporate setting, since Linux in the glass house is already well established.

Linux Is Hard to Use/Learn/Operate

Does anyone remember moving from DOS to Windows 3.1 or from Windows 3.1 to Windows 95? If people can make those jumps, then Windows 2000/XP to KDE should not be nearly as hard.

Here is a simple example: How hard would it be for an end user to adapt from the look and feel of MS Windows 2000 (shown in Figure 1 running as a VMWare guest on my laptop) to the desktop shown in Figure 2, which is KDE 3.2.3 running under Mandrake 10.1?

The user interface paradigm is the same: click to do things, such as opening up menus that contain actions. Click on icons on either desktops or task bars to start specific things right away. See the status of things on the right hand of the task bar; it's just not that different. Nothing like MS Windows 3.1 to MS Windows 95, and that jump was called revolutionary, not "too hard to use because it is a little different."

Further, my daughter has been using Linux and Windows interchangeably since she was seven. Are the purveyors of this one saying that everyone suddenly got stupid...right after they converted to Windows 2000 or XP?

Linux Is Hard to Install

In a corporate setting how easy or hard an OS is to install doesn't matter that much. It's usually centrally installed, then imaged out to every-one else. But that's good news for MS Windows more than it is for Linux, because no current major distribution is harder to install than MS Windows and, several, notably Xandros and Linspire, are much easier. Last time I installed Linspire it asked me four questions, and one was about what I wanted to call this computer...

Related to this point: Linux has all sorts of centralized management tools available too. Once it's out there on all the desktops, you can exert central control every bit as well as you can in an MS Windows world. Some vendors understand that you'll have a mixed environment and sell centralized management tools that will work on both MS Windows and Linux desktop systems.

Linux Is Too Complicated

Again we are getting our intelligence insulted. Linux has one major complication over MS Windows: choice. Someone has to decide things like which distribution you are going to use, which browser or browsers you are going to preinstall, which office suite you are going to use, and on and on. You make these choices not because you have to but because standardization decreases your training and support costs to some degree.

This "complicated" argument is disingenuous - MS Windows is hardly a simple environment; it has over 20 million lines of code in it. The minute you try to stray off the beaten path, you find yourself in the middle of trying to figure out the MS Windows registry or something, very complicated. Very scary!

My daughter never had any trouble running Linux at all, way before she was the 11-year-old computer literate she is today. For what it's worth, her first distribution was Debian, clearly more "complicated" than the current Fedora or Mandrake or SUSE.

I could go on, but let's move on. Identify the FUD that has taken root at your company. Develop a reasoned response to it. Document it on a standards-compliant Web site. And move on.

Getting People to Move Off MS Windows Desktops

Yes, this would apply to moving to an iBook every bit as much as moving to a Linux desktop. Yes, this is a hint to our corporate standards folks. Actually, it's interesting to note how many attendees at LinuxWorld had iBooks. But I digress.

There are many reasons people won't want to make the switch to a Linux desktop. Fear of change is no doubt the biggest. But there are others:

  1. Someone may have bought into the FUD.
  2. They may fear that this will make them lose their job: after all, they already know MS Windows.
  3. Some people like MS Windows. It's "safe." No one would ever question you if you said "I lost my presentation when MS Windows crashed" or "I can't help you on the document; IT is rebuilding my PC after that last virus outbreak."
  4. Some people like MS Windows because it's all integrated, works together, has eye candy they like, and removes some of the choices they'd rather not think about in the first place.
How do you get them to move? It's not enough to tell them that the company just wants to go that way so they can save themselves downtime from viruses or money in year two after the conver- sion. Some will jump just because they like change; you always have the "early adopters." Some will move because they want the company to succeed. But many will resist. How do you get them to move?
  1. There is always the ever-popular corporate mandate. Sun, IBM, Novell, and others have these - great when you are doing this from the top down. Related to that is the government mandates: see France, Germany, and China for examples of these.
  2. Middle management can use bonuses, if they have them, to motivate the behaviors they want to see: pay folks to convert to Linux and structure it so that the sooner they convert, the better the bonus. Watch what happens.
  3. Training - Web-based, instructor-led, or one-on-one - make available things like Linux desktop certifications from places like the Linux Professional Institute (www.lpi.org). Folks comfortable with the desktop for things like "How do I edit this document?" or "Send that note to Sheila" will be much more comfortable with the change.
  4. Make sure your MS Windows system administrators get trained as Linux admins too. Some won't like it. Some will fight you if you try to make them work on Linux. I've had it happen both ways. But the job assurance they will get from being cross-trained will help ensure your success. Even the ones fighting the change said they were glad they would have the skills they needed to stay in the new environment.
  5. You could always tell them that you know about this seven year-old who can do it...

Risk Averse?

All change, regardless of what it is, comes with risk. Standing still has its own risks too. There is no such thing as a risk-free environment. Leaving that aside for the moment, it's true that some may perceive that moving to a Linux desktop contains more risk than staying put in the current environment. Total cost of ownership issues aside (as that really would be an entire book, not an article), how do you find out what the risks are in your particular environment? One way is to commission a skunkworks project. Find the group in your shop that has always liked to march to the beat of a different drummer; every shop of any size usually has one of these. In a technology company you probably could stack them like firewood.

The mission of the Linux desktop skunkworks would be simple: find out what it would take to run a Linux desktop here in our company. We're not like everyone else (we all always think that), so what works and what doesn't? What would need to be changed where? I like lists, so here is another one:

  1. Document everything on a standards-compliant Web site. By that I mean it would be very silly to publish your Linux desktop data on a Web server using vendor-specific HTML so that you could not read your own document unless you were on a specific browser. Open Standards for Web content are easy, work well, and interoperate. Use them! (See www.w3c.org for details.) Tools like OpenOffice and Mozilla are pretty good at writing content in a standards-compliant way.
  2. Document the Web applications that are working already.
  3. Document the Web applications that are not working. Try a few different browsers to be sure that you really did paint yourself into a corner here; sometimes Konqueror will work when Mozilla or Opera won't. Let your Web team know as soon as you can about the apps that are not standardized enough to work cross-platform. Most Web admins I've met hate it when their Web pages don't work and will work with you to fix it. If they say "just use IE," consider this: any Web application that uses only one browser is not a Web application; it's a client/server application. Are you moving toward client/server or away from it at your shop?
  4. Install and play around with the technologies that let you run MS Windows applications or even MS Windows under Linux. I'm speaking here of things like WINE, Codeweavers Crossover, VMWare, Win4Lin, and so on.
  5. Do you have MS Windows servers running Citrix or MS's WTS? Try TSClient and rdesktop. Have a look at VNC. Understand those things that don't work natively and do work in a thin-client mode instead.
  6. Put all your how-tos on your skunkworks Web site so that your early adopters can get a head start and when you start the bigger move, you have your doc mostly prewritten.
  7. Look at collaboration technologies as a way of allowing your desktop users to help each other, such as Blogs, Listserv (www.lsoft.com/default.asp), and MajorDomo (www.greatcircle.com/majordomo) mailing lists and collaboration Web sites like Citadel (http://uncensored.citadel.org/citadel/), maybe even an internal NNTP server. There are a ton of free, open solutions out there to allow your folks to better work together. Pick the one that works best in your environment and with your end-user set. Maybe you already have a commercial collaboration solution in-house? Does it work with Linux? Give it a whirl.
  8. How do you centrally manage your current desktop environment (or do you?), and how will you extend that to cover Linux? Test it out. There are several vendors out there with management solutions that will run across multiple platforms.
  9. Anti-virus - same as above. Several vendors have solutions that run on more than one desktop: does your vendor? If they don't, ask them when they will. It may be sooner than you think. Even though Linux does not get most MS Windows-based viruses and worms, it can harbor them on network disk shares. You'll want to think about extending whatever you do to cover the new desktop. Have your skunkworks folks give the A/V software a whirl and see how it works.
How much of the above you have to do will also have to do with where the push to desktop Linux is coming from at your company. If you have a top-down or middle-out initiative, you're more likely to formalize all this. If this is coming into your place the same way Linux in the data center did most places - a grassroots or bottom-up phenomena - then you have most of the same things happening, just not as formally. You still have to figure out how to work in your environment, even if all you are doing is documenting it on the Apache server running on your laptop for all your friends and department mates to use. Hang on to that data; the corporate IT people may want to use it later.

Training and Certification of Desktop Users to Diffuse Anxiety

Microsoft has a certification for users of their application called MOS for Microsoft Office Specialist (formerly Microsoft Office User Specialist). Similar things are in existence or being developed for Linux desktops. One easy way to reduce the FUD of converting to a Linux desktop is to have some key people take the Linux end-user certifications. Once they see that typing a document into something like OpenOffice is not all that different from MS Word or WordPerfect, then some of the fear of the unknown will be diffused.

I would not make such a program mandatory. There should be plenty of folks who want to do it, and they will also be the ones who will help out the others who convert later in the process. You set up that open standards-based collaboration site, right?

This is part of how the desktop conversion process stays technical rather than emotional.

It will happen that "computer religion" will creep in, no matter what you do. Just like nasty rumors, efforts should be made to stomp it out whenever it rears its ugly head. The simple fact is that converting to a Linux desktop now or in the near future doesn't mean all the people who were using MS Windows or some other desktop were somehow bad or stupid. You can't win emotional arguments here. Converting to Linux desktops was a business decision, and your presentation of the migration should be made at that level. Maybe you shouldn't remind those reluctant to convert about the seven-year old; they may think you are trying to insult them.

Choices: Vendors and Standards

There you are. You need to get started on this whole corporate desktop Linux thing. You need to make some decisions, based on your company's situation, to get to the setup that will work best in your environment. The skunkworks project probably helped you figure out some parameters for what it will take to get this done, but there are some others that are more arbitrary. For example, you decided that to get along with application "A" you need to run it in WINE until the application support folks can get the Web interface standardized enough to run it under any browser (you are getting rid of all that client/server code in your Web apps, right?). But that did not tell you anything about which distribution you need to run, or which hardware vendor.

On the other hand, you probably already have a hardware vendor in-house. In fact, if your shop has been around for more than a couple of years, you probably have had three or four vendors come through. Some folks have Dells and others have Compaq, and the laptops are all Toshibas or IBMs or something. The first thing you need to ask your hardware supplier is what desktop class hardware they support Linux on. Last year this would have caused consternation and much gnashing of teeth. I even had one major PC vendor tell me it was illegal for them to sell anything other than MS Windows on their laptops. They were very surprised when I pointed out all the scofflaws on the Internet that sell Linux-only laptops.

This year is a different story. HP has their new NX series line with SUSE Linux. Dell has a deal out there on their D600 with Xandros (although that's probably not direct from Dell). IBM has a Linux laptop now. And these guys are only now joining the folks that have had them for years, people like sub300.com and emperorlinux.com, to name two.

Knowing your hardware vendor's support is very useful when taking in data from your test team about what is working well in other ways. Let's say you're using MS Exchange for e-mail in whole or in part. A likely result of that will be that you'll want to use Ximian Connector to allow your Linux desktop folks to interact with your MS Outlook users, at least for the transition period. If everyone is not going to Linux, chances are there will always be Outlook folks lurking about your network, sending badly formed MIME attachments in their e-mail. That is also another article...

If you're going to use Ximian Connector right now, you're limited to a distribution that has it already. Even though the connector is open source now, not all the vendors have it ported at this time. When I looked at Mandrake 10.1 beta 1 last week as part of our skunkworks project, I didn't see the Exchange Connector included with their port of Evolution yet. Novell Groupwise was though.

Now maybe it will be in sharper focus for you: your hardware vendor likes one or the other distro and your internal applications need one of a limited set of distros...you triage this all down till you have what you need for your environment.

Provisioning and Patching

To succeed with a corporate Linux desktop, it has to do more than have a pretty face, interoperate with MS Windows-based servers, be less expensive over the long term, and easy to use. Linux also has to be easy to deploy and patch. All the Windows patching and virus outbreaks have taught us that in a corporate desktop environment, centralized management is a must. With this in mind, we'll look at how we would go about creating a standard image, deploying it, and then maintaining it "down the road." The good news here is that this is a relatively mature area for Linux. If you have this all deployed for MS Windows, your vendor may already have the Linux version either out or in the pipe. If not, you will have several more choices to make here. You did want more choices, right?

Windows Terminal Server Access/VNC

There are some applications that just don't work well inside of Linux yet, sad but true. Often the Web interface is somewhat workable, but does not yet have all the capabilities of the native MS Windows client. While you could run the client under WINE, another way is via a local terminal server. We've tested using this with great results, as Linux has a terrific MS Windows Terminal Server client available: rdesktop and its GUI front-end TSClient. Most distros ship both, but some don't have the TSClient. No seats on your WTS or Citrix? You can also get an MS Windows system and load up the VNC server and then access it from your VNC client on your Linux desktop. This is single-threaded, one-at-a-time type access, but if you only have a few legacy apps you need to run over there, it might work out really well.

Teleconferencing

One of the things we started using Linux for right away was teleconferencing using GnomeMeeting. While we use it GnomeMeeting to GnomeMeeting, it's also possible to use it GnomeMeeting to MS NetMeeting. It will be just video and sound: no proprietary extensions interoperate. If you need white boards, chats, and shared desktops, you have to use things like KDE's desktop sharing or IM to augment what the GnomeMeeting to MS NetMeeting interoperate brings to the table. There are more Linux IM tools than you can shake a stick at: I use Kopete or GAIM. Mostly GAIM, because it spell checks.

Document Formats

We all did it; we all drank the MS Office Kool-Aid, and now we locked ourselves into this closed document format. But unless you have been sleeping under a rock, you'll have heard about StarOffice and OpenOffice. The spreadsheets macro conversion stuff is still a work in progress, but as long as you have the same fonts everywhere, your .doc stuff will work as well or better under Star/OpenOffice. Better? How can that be? I have no idea how they did it, but I have had cases where I could not open a .doc-formatted document under MS Office and could open it with Star/OpenOffice. Then I saved it in .doc and suddenly I could read it un-der MS Office again. This was on Windows: one of the very nice things about projects like Star/OpenOffice is that they will run on all sorts of platforms. I just wish they had a "Show Codes" feature like WordPerfect.

Then there is the option of running MS Office under CrossOver Office. Now you have Excel under Linux, and it works pretty well too. Since I want to get away from closed document formats, I want to get away from .doc and .xls and so I only use tools like this to get at legacy data. Everything new I do in Star/OpenOffice. Other apps, like Koffice, are able to read and write to their formats, so I have no application lock-in there. Star/OpenOffice can save to .pdf, which I am certain has a reader on most every platform. When I send someone a document and they can't read it, I next send them a .pdf of it and have never had a case where that did not work.

As it turns out, the closed document format problem is actually pretty easy to solve - this time. This doesn't mean you should leave it there. There are risks involved, and people are trying to copyright file formats and XML Schemas and whatnot. Get out while you have a chance; the next lock-in may be worse.

Pulling It All Together

At this point you've looked at your apps and tested them in a variety of ways. You've talked to your vendors. You know how you are going to approach your environment and you've whittled down the many choices of Linux to the way you are going to go after it: which vendor you will have in place for hardware, distribution, provisioning, backup, DR, and so on. You know whose neck you will wring if something doesn't work, if you have the type of shop that likes to have necks extended for such things. You know which applications are true Web apps and which need some work.

You know that if you made a wrong turn, with Linux you can work your way out of it. You are not locked in to anything. The corporate Linux desktop: it's here.

More Stories By Steve Carl

Steve Carl is the manager of R&D Open Systems Support for BMC Software, Inc. His team supports over 50 different versions of Unix, Microsoft Windows from NT 3.51 on, and a wide variety of Linux distributions across many, many platforms. He's a member of the IT Architectural Council, a group charged with positioning BMC Software for long-term technological changes. Related to that, he is currently investigating the feasibility of Linux on the desktop for BMC Software's large user community.

Comments (4) View Comments

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.


Most Recent Comments
Brian Masinick 11/30/04 09:25:26 PM EST

Linux has never been truly difficult to use, but it sprang up from UNIX roots, so it had a reputation for being different and unusual, compared to what most computer desktop users are used to using. In that sense, I suppose it is still true today, but from the first time I used Linux back in 1995 until the present time, I would not characterize it as being difficult, just different. In 1995, there were already window managers available on Linux that made every day computer usage reasonable. Back then, the system installation was beyond the caaual user, the graphical environment had to be configured by someone what knew what they were doing, but suppose you had your software installed and administered by a professional system administrator? I'd maintain that even back then, there was workable, usable software. It just wasn't intuitive for the casual user to understand or set up on their own.

Today, however, that's not even true any more. These days, you can turn on your computer system and begin running whatever happens to be on it, then you can insert a CD containing Linux software, reboot your system, and have it come up running really easy to understand Linux software that even looks and feels a lot like whatever you've just been running. Usability is here. Applications still lack some of the depth of the commercial stuff, but even so, there are at least a few usable, free or inexpensive applications in every major desktop category today.

I would assert that Linux has arrived. All we need now is a growing critical mass of development, support, and wide scale adoption so that hardware support continues to improve and so that people's favorite application can be ported to the platform.

Today I use Linux nearly all the time. The only time I use other software is when I am at work or using someone else's system. I run either Linux or BSD software on my own systems. I can run office suites, Web browsers, and text editors that can seamlessly communicate with people in ordinary offices with no trouble at all.

pjg87 11/29/04 02:38:41 PM EST

One of the things we started using Linux for right away was teleconferencing using GnomeMeeting. While we use it GnomeMeeting to GnomeMeeting, it's also possible to use it GnomeMeeting to MS NetMeeting.

Wow I'd not realized there was interoperability between MS NetMeeting and GnomeMeeting.

just_an_it_guy 11/24/04 06:08:48 AM EST

Linux is growing. Customers want a server alternative and we have been seeing that market grow for Linux the past few years. Enterprises want a alternate desktop. They want tools to manage their Linux or Cross platform Networks. Enterprises want products to manage these new Linux desktops. Novell is going to give the community all of those things customers want. Sure not everyone will buy Novell products, but companies that want quick drop in solutions that work will end up with Novell products. Companies that want to run Cross platform networks where everything is spelled out will want Novell.

DesktopPenguin 11/08/04 05:53:50 AM EST

Linux has one major complication over MS Windows: choice

haha, great line!

You are right, corporate dekstops are certain to start running Linux at an ever-increasing rate.