Slashdot Mirror
Linux in the Workplace
Linux has failed to catch on among ordinary users because PC manufacturers have been prevented from offering dual-boot systems. Not only is partitioning and dual booting a little tricky, the OS CD that comes with a new PC is usually just a system restore, making it difficult to configure a dual-boot environment without messing up factory settings and file systems. Learning Linux has become an all-or-nothing proposition; in many cases the new user has to wipe Windows off his system for good or rely on a second machine just to get started.
The irony is that Linux has never been more user friendly, and the latest KDE desktop on my gentoo box is slicker, faster and easier to use than XP. It is becoming easier to be productive on Linux, and while university students have already discovered this, corporate IT departments who support a large number of Windows-only commercial applications tend to view open source solutions as a time burden (these are the same IT departments whose days are consumed with applying Windows patches or verifying license compliance).
The book Linux in the Workplace shows the ordinary user who has never laid eyes on Linux how to perform everyday office tasks. The book assumes that the user has a machine with Linux already installed and successfully configured. This book (which is more of an introduction to the KDE desktop than Linux itself) is easy and fun to read, and has lots of screenshots. Slashdotters might find this book a bit too basic, but it's the kind of book that a technophobe spouse or child or parent might love (and could very well appear under Christmas trees right beside the new Linux PC).
This book devotes a chapter each to talking about Open Office, Gimp, Konqueror, personal information managers, and various KDE office and email applications. Most of them are part of the KDE window manager or installed by default. This book walks a thin line between being too superficial for daily use and dwelling too much on the technical details. There are better books on The GIMP or OpenOffice, for example, but still it is nice to have introductory chapters in a single book. I found a few useful tidbits on controlling file associations, xscanimage, screen capturing and ark archiver. The book is not without a sense of humor. In a useful section on creating a GPG key, the book says "your passphrase should be rude or embarrassing ... using a naughty passphrase will remind you not to type it where others can see."
This book began with the mission to bring a simplified approach to Linux. By definition, it must exclude certain topics, either by design or because an application was not yet mature when the book was being written. The book scrupulously avoids a discussion of server applications like apache (which makes a certain sense), but it would have been nice to have a section on Evolution or mplayer (yes, a media player is an indispensable application for the bored employee) or ssh, cd burning programs, browser plugins, mozilla, crossover, irc or ftp clients.
Surprisingly, the book contains almost nothing about printing or how to install or upgrade applications. Because the book is intended for a newbie user, not a sys admin, it recommends talking to your network administrator about that. Cop out? Perhaps. But even the unskilled non-root user will have to install apps once in a while; the book would have been much better with a section on rpm managers and compiling programs from scratch.
From a sys admin's point of view, I would have liked to see a case study of an office that had actually made the switch. What problems did it encounter? How did the switch change business processes? What applications required the most time and energy for support? How did a Linux-only office manage domain authentication or interoperability? What system management tools made administration easier in a heterogenous environment?
The book raises an epistemological question about the best way to learn a new technology. Will a user who has never really performed tasks as root be able to leverage the freedom and power offered by open source? Will a user truly be comfortable with an operating system without first having experienced the agony of a bad install or frantically scouring the newsgroups for help ? This book presumes that a learner needs to be able to use normal applications before being ready to handle the admin stuff. The problem with that approach is that it depends on IT staff being near and ready to do some hand-holding. But Linux may emerge in the workplace not as a result of IT's enthusiasm but because of ordinary workers' exasperation with uniform proprietary solutions imposed by these IT departments.
In summary: A useful and friendly KDE-centric introduction to Linux for nongeeks. The lack of system administration material makes it probably too basic for slashdotters.
Also recommended:
RUTE Users' Tutorial and Exposition
A Practical Guide to Red Hat Linux 8 by Mark G. Sobell (not yet published)
Robert Nagle is a technical writer, trainer and Linux aficionado in Houston, Texas. You can purchase Linux in the Workplace from bn.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.
As a network admin we deploy Linux servers (Debian, thank-you-very-much) because they simply kick ass. Sendmail, Squid, Apache, FreeSwan, iptables, etc., all blow away similar products put out by Microsoft. I wish more IT people would seriously consider Linux at least for server duties.
"Maybe geeks and slashdotters understand why Linux is so appealing"
No, geeks and slashdotter think Linux is appealing. It isn't a fact that it is, some people just think it is.
This could be a book that creates its own market. As more people buy the book, more people run Linux in the workplace, thus making demand for the book rise.
I would make a 3.Profit! joke here, but I think an evil laugh is more on order. [evil_laugh]Muahahaha[/evil_laugh]
I don't keep a lid on my coffee so when I walk around I look busy -me
I've always been surprised and even disappointed at my friends' lack of curiosity about Linux.
Linux on the desktop sucks for one reason: fragmentation.
If there wasn't a bunch of if distribution == x && graphical environment == y in the HOWTOs, we'd be much further along right now.
Even with all that cruft, the one itel holding Linux back is the file system requirements. Every mainstream consumer desktop sold these days has a hard drive installed with a 100% NTFS partition. People don't want to screw with boot managers and people don't want to screw with repartitioning.
If someone created a *free* distribution that could be installed *through* Windows on the local NTFS partition (with appropriate *free* boot manager), then we'd have a much larger installed base. Developers could start coding on Linux and distributing it with their applications.
But then there is the GPL/binary module hassle. it will never work. I'm waiting for Apple to port over to x86. I predict 2004.
Life is the leading cause of death in America.
I think that gentle introductions to Linux for Win users are a Good Thing. Someone who decides to buy this book probably already has some level of interest in Linux, and is looking to see if the switch can be done with little-to-no pain.
The real challenge, is getting people to that point. One approach is to have lots of stories published in the mass media that talk about how easy/efficient Linux is. The challenge there is not to raise expectations too high. If someone expects to be able to sit down in front of their computer, put a Linux CD in, click "ok" a few times, and be up and running, doing everything they had been able to do in Win, they'll be disappointed, and are likely to give up. Non-geeks aren't motivated to hack around for a while. They want to use their new tool.
People need appropriate motivation. As an example, speech recognition software is more likely to be successful when the user has a strong motivation to work through the early hassles. People with RSIs or other physical constraints are more likely to become successful speech reco users than are ablebodied people.
So, the challenge is to motivate people to try it without raising expectations too high. I'm not sure what the answers are, but although this type of book is a good step, more needs to be done.
Linux has failed to catch on among ordinary users because PC manufacturers have been prevented from offering dual-boot systems.
:)
The reason I've seen a few geek friends try out Linux then walk away disgusted (possibly forever) is hardware support. Sure, with modern distros almost all common hardware available is supported, but in a lot of cases it requires a kernel recompile, some config file changes, sometimes even low-level stuff like probing around to find out an IRQ setting.
Compare this to Windows. Not only does almost every piece of hardware come with a driver, most people are comfortable with the driver install process (and the ones that aren't usually have a family member or friend that is willing to do it).
Too bad Linux kernel & distro developers can't create a kernel standard for common release, and just put a stake in the ground and say "Here's Linux 2003. Any certified standard common pre-compiled driver module dated 2003 or older will work with this years Linux." No recompiling the damn kernel. Then of course there'd have to be a very standard common driver installation program... And the rest of us who want to recompile our kernels can still do it if we want.
Eh, but what do I know. I run BeOS and a driver is typically one file that I drop into a folder and usually begins running immediately.
_______
2B1ASK1
Don't be. It sounds to me like you have a normal and well rounded set of friends. Good for you.
These people think of their computers as a tool just to get work done
That's me. I may be a geek, but I like having a life. A computer is just a tool, like a hammer, or saw, and so is all the software on a computer. I use whatever tool works. It was difficult to make the switch to Linux (it had been 10 years since I had been doing serious programming!), but I did it. And Linux, as a tool, works much better than Windows. It's like using a high priced hammer that absorbs some of the impact instead of a hammer that breaks in two several times a day.
In a way I don't care what OS my box runs. If it works and does the job, it's what I want. Unfortunately, I even had problems with Win2k -- which I called Win69, since it went down on me so often (not as often as Win9x, but still way too much).
You make linux out to be some fanatical cult thats the best thing in the world
If you care about an OS that works, that doesn't crash, that isn't full of security holes, that doesn't cost an arm and leg to upgrade, that doesn't act as a platform for an office suite that costs (literally) hundreds of dollars, then it is. A lawyer friend of mine just convinced his wife to go for Linux on her new system. How? She looked at the pricetag and realized by the time she got a system with WinXP, Office (a full or almost-full version), a finance program, and the one or two other things she needed (a total of WinXP plus something like 3-4 products), she would spend $1,000 on SOFTWARE alone! She's using Open Office now.
people don't care about such trivial things
I'm not clear what you're referring to as trivial. (The noun substitute "things" does not have a clear reference.) If it's so trivial, why are you responding? It isn't trivial when companies like Dreamworks, Merril-Lynch, BP, and many other huge companies decide that Windows is costing too much or not doing the job and switch to a system that doesn't crash and doesn't result in large licensing fees being extorted from them on a regular basis.
But each to his own. If you want to pay more and get less (except for pretty bells and whistles), then, as Sirius Cybernetics said, "share and enjoy."
Everyone talks about KDE/Gnome as the reason to switch to linux, linux is the kernel and drivers for the hardware. Users want applications and a nice looking desktop, eyecandy.
Personally, I like XP as my gui, and Linux as my server and extension to my workstation. I use my linux box as resources for my windows box. Mount shares, Run services, shell with command line tools, keep tasks running in the background while I play video games on my windose box. (Keep IRC open in a shell with irssi)
I really dont understand why people dont use the best of each platform. Have the best of both worlds, the power of applications (and GNU utils/commands) on a *nix box(bsd or linux) and the anti-aliased fonts/games/apps of windows. (Sounds like OSX, doesnt it...)
But if you only have 1 Box, XP+Cygwin seems a better option for now. KDE/Gnome/Openbox are still lacking in areas, and windows programs wont run native. Repeat, im not bad mouthing linux, linux is a great OS. The Gui KDE, and Back end Xserver is missing features, 3D features, Anti-aliasing, advanced hardware features. But the command line gnu tools, and opensource applications are great. I'm sure with time, linux will be the better choice for a full time os, but gui and applications support needs to be there.
-Brook
-
Blackbox 4 windows an alternative to litestep.
while humorous, the analogy isnt worth any more than a chuckle.
I hear what you're saying, but the members of this technical forum are far more technical than the average Windows users, especially at the office. Joe Blow has 5 years of experience with MS Word, Excel and Power-Point and wants the same apps at home- that's all there is to it. To make that work he needs to be able to call Dell, order a new PC with Red-Hat and Office 10 for Linux. If it's anymore complicated than that he will go with Windows every time.
It's amazing how few people have switched to Mac's even with this capability. I bought a new iMac just for grins, and with 10.2 and the Cisco VPN client, I can dial up my office and fire up Outlook to the Exchange server no problem, as well as have CRUD access to the docs on the network.
Gotta get Office on Linux without any adapters, emulators or dual-boots. The average user, (who is keeping the lights on in Redmond), isn't as fast as this community.
I wonder why people keep repeating the mantra "Linux is more stable than Windows". The operating system might be more stable, but it doesn't mean its applications are. And unfortunately, most of them aren't.
The problem with Linux is that it's open source. It makes everybody think they can program. So the cycle goes like this: Jill writes a small program for herself. After a while, adding and changing (and because of the famous "release early, release often"), she puts it online, and says "The code is a mess - I'll get back to changing it as soon as I can". After a while people start using it, until some feature is missing. So little Joe decides he can program. He opens the source code in his favourite editor, and behold - it's totally unreadable. What does Joe do? Instead of working with the original author, he decides that he's good enough to write his own version. And so on.
The problem is that a normal person needs a certain tool. He looks online for a "financial calculator". He finds 127 different versions, each with a version number 0.01alpha, and then he has two options: First, install Windows and buy software from a company he's heard about, or wasting his day compiling these 127 versions, just to find out none of them actually works (why compiling and not installing? oh, they were packaged under Debian, and he's got RedHat with a different glibc version).
We have at work Windows XP and Linux. Each of them is good for something, but for general use (which includes debugging, by the way), Windows is simply better. What do you use on Linux? gdb? ddd? gvd? Compare that to Microsoft's debuggers, and you'll see the difference.
It's not there yet.
All of my servers are Linux, and they do great. In fact, I'll turn 1 year uptime on them next week. However, I don't think Linux is ready to be on my desk at work.
Every once in a while, I get this urge to try to dump Windows altogether. I've tried it several times now, but I keep coming back to Windows because of apps like Photoshop and Trillian, and the solid UI. There are X equivalents of most apps, yes, but they just aren't the same, and I'm not as productive with these as I am with the Windows products.
Most Linux desktop apps have not been very stable for me either, and what's worse, they don't FEEL stable. MS Windows has a very solid, polished feel to it. They've dumped tons of money and hours into useability, and they have alot to show for it. I think Linux will get there (its come a LONG way already), but for me, its not there yet.
Last time I ran a Linux desktop, someone asked "why are you running Linux instead of Windows?". I really couldn't come up with a valid answer for him, other than "I just want to!" or "I hate Microsoft!".
The bottomline is, right now I'm the most productive when working from an SSH session on my Windows desktop.
I'm a long time NT guy (NT 4.0 MCSE, Citrix Metaframe 1.8 CCA), and while I had played with Linux 5 years ago, next saw any appeal on a desktop. My first experiment with Linux was in 1997. About 40 hours later, I had it all configured to my tastes, everything worked, and I wondered what I was going to do. All it was was a slightly inferior to the NT 4.0 desktop that I was using at the time. Forget games, I was an NT guy, we had no games either.
At my business, we deploy on PHP 4 + PostgreSQL, so we have Linux database servers and OpenBSD webservers. Our first Linux web server in 18 months just came online, we got sick of security issues.
I currently use a Powerbook w/ OS X for my desktop, I'm extremely happy. When we were playing with Redhat 8 to install the test box, we did one install as a workstation for fun.
It was distinctly less ugly than I remembered Linux desktops, and was pretty equivalent to a Windows desktop (though it can't touch Aqua). However, when I tried to install Phoenix, I ran into dependancy problems because I hadn't installed Mozilla first (I was going to run Phoenix). When I created a "launcher" I couldn't get it to show up on the desktop until relogging in, etc., etc.
If I was a grunt office user, I could be trained to work in there instead of Windows. Someone else would create all my icons, etc. For Sysadmining, I have no problem playing in Linux, its easily to configure, etc. However, as a "power user" I was frustrated, and wanted nothing to do with the box.
I find OS X + Powerbook makes me EXTREMELY productive. Redhat + GNOME + KDE + Blue Curve was too frustrating. It's "looking" better, but it isn't better.
Look, there are plenty of times that I get confused in the Mac GUI because it isn't Windows. I can usually figure it out, and the result tends to make more sense than Microsoft's version.
With my Powerbook, I plug a second monitor in and the dock/menu bar slide over. When I disconnect the monitor, I'm back to one monitor. BBEdit has configuration options for working with two monitors, very nice. With my Windows laptop, I had to shut down to undock b/c of the PCI video card to get the second monitor. How would Linux handle that?
As a result, Apple go the check. Switching was only a few thousand, and I'm more productive. Knock off one extra project and its paid for itself. Give me another two weeks. Linux... sorry, its not there yet.
Alex
Printing can be a huge problem in the unix world.
/etc/printcap. It's got a beautiful web-based GUI and supports IPP.
Have you ever used CUPS? It is the easiest Linux print system I have ever set up. No more messing with
I control the time!
Would you also be disappointed in my lack of curiousity in Microsoft Windows ?? It's come a long way since the bad-old-days of Windows 3.1. That's the time frame when I purchased my first PC and installed Slackware 1.1 (0.99pl14 kernel). Previously, I used a MacLC2 (dual boot, MacOS6 and MacOS7), and I logged into to various unix boxes for "real work". Before that, I had an Apple2, and again, I logged into BBSs and unix boxes for email, newsgroups, and chat.
I kept the Mac for many years and used it for word processing and graphics (bought one of the few monitors at the time which had two video inputs). MS Word 4.0, MacPaint 1.1, SuperPaint 3.0, Canvas 2.1 were getting pretty old, but they still worked great and did everything I needed. My old Mac has a 50 MHz speed-up card, and those old apps ran great. Likewise, I could do almost all unix-oriented tasks on the linux box, including email, usenet news, and later surfing the web. Linux (and related apps) has grown and grown, and the PC hardware has remained cheap (unlike trying to upgrade the mac). A couple years ago, I took the plunge and finally started using the GIMP, which replaced my last major hold-out on the Mac side.
Over the years, there's been 2 win32-only CAD apps I've needed. At times I had dual-boot, but eventually I purchased vmware and I really like the repeatable resume. I can finally not have to fiddle with windows.... I just set it up once and every time I start that virtual machine I get exactly the same working win32 system with my one CAD app installed.
I saw WinXP in the store not long ago. They've certainly made it pretty. It also looks like Win2k and WinXP are real operating systems with compatible apps and drivers (I was quite unimpressed with NT 3.51 and 4.0).
My linux setup works. I know how to use it. I have a set of apps that run great and do just about everything I need. I've got all my special apps in /usr/local and ~/bin, so backup and migrating to newer distros are easy.
I know there's LOTS of neat new apps for Windows that don't exist for Linux. I know the modern versions of Windows have become much better.
But I don't really care. What I have works, and until there's some really compelling reason to consider Microsoft again (that isn't easily solved by a repeatable-resume vmware virtual machine), I just want to leave well enough alone. It's certainly not broken, so why fix it ???
I can identify with your Windows-based friends who are quite happy with their computing paragigms and therefore aren't really curious about Linux. I can't see how yet-another-book is going to "help". If everything is working great and there's no need for anything new, then what is the "problem" that needs to be "solved" ?
PJRC: Electronic Projects, 8051 Microcontroller Tools