Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
×
Linux Business

Free Software's Star to Rise During US Recession? 265

robvasquez: "Surfing the web today, I see that RedHat has broke even and that Corel has shown a surprising profit (they are they still considered a linux company, right?), so I'm seeing Linux companies out of the red! Perhaps this 'recession', whether we are starting one or even going in to one, is what we need to popularize free software. Think about it: with companies laying people off and cheapening up, whats better than free software? They could lay off 5,000 people, OR quit shelling out that much in MS licenses and pay salaries!" It's a nice thought, and a good idea for established companies to look into, but quite a few Linux companies have been hit just as hard during the recent economic crunch. I guess only time will tell if the use of Free Software is as much an economic advantage as people have been making it out to be, and now is as good of a time to test this claim as any.
This discussion has been archived. No new comments can be posted.

Free Software's Star to Rise During US Recession?

Comments Filter:
  • by Anonymous Coward
    First of all, just because Linux systems are free to purchase does not necessarily reduce the cost of ownership. Granted, I agree that it would be cheaper than running everything from Microsoft.

    However, large corporations use hundreds of file and printer servers for basic network connectivity. Then factor in the millions spent on highend systems like Sun Hardware for mission critical systems (data warehousing, customer service, etc).

    Suppose a lot of the above is replaced with Linux systems were it currently makes sense. The only folks in my company that could handle supporting these systems would be the Sun guru's. That leaves out the following people. Helpdesks, LAN Centers, and Desktop Tech's. That's about 300+ techies who don't even know what Linux is.

    Where in the heck are we to find enough people and pay them industry rates? Sure we can buy contracts with RedHat, etc for support but they aren't going to handle every single little problem. We would have to train a whole mess of tech's to get them up-to-speed on how a real operating system works. Most of our IT Staff are clueless, mindless, idiots who don't know a thing about Unix or Unix like systems.

    According to my count, there are approximately 50 IT people enterprise wide that really know what's going on and could easily handle supporting Linux. Sadly, most of our IT Staff is lame.... It's hard enough to just get people, let alone quality people. The top 50 I mentioned are mostly Network Engineers, Programmers, and Unix guru's. But the rest of IT is basic support staff.

    "What do you mean just Secure Shell into it and stop the database daemon and restart it? Can't we just reboot it?"

    The cost of training everyone is ridiculous... We need more of a grass roots movement to get the general IT staff up-to-speed on Linux. Unfortunately the majority of IT workers are in IT because they were told it's a good career. They are not all geeks. The true geeks would be doing this job for no pay if they didn't have to worry about putting food on the table.
  • by Anonymous Coward on Tuesday April 03, 2001 @09:20AM (#318038)
    Even if the software is "free", the implementation and maintainence of said software is where a lot of the costs come into play. The talent pool for MS software is broader than that of Linux (or other "free" applications), so it would seem to me that hiring an equivalently competant staff of Linux gurus over MS gurus would cost a bit more. The costs of running a successful IT department aren't only those of liscencing, but of keeping the IT staff trained and up to date.
  • ...spoken like a true noncreative non-producing type person who _has_ nothing to produce or contribute. Spoke like a stockbroker.

    *ahem* [airwindows.com]

    Speaking as someone who _does_ produce things, you can take free software away from me when you pry my cold dead fingers off it- when I write something it's _my_ choice how to license it, not yours, and it's simply comical to see you insisting that free software will go away. Not a freaking chance, Sparky. What have _you_ written besides Slashdot comments?

  • by Tim ( 686 )
    "Linux doesn't save any money to a IT shop due to licensing fees becuase windows licesne fees are not that much. However linux can save a lot of money if the machines are set and forget. administrators are the expensive part."

    Yeah, admins are expensive, but at thousands of dollars for a 5 user license pack for Win 2000, the difference starts to diminish -- remember, we're not only talking about the big iron in the back room (which, frankly, is likely to be running a *NIX anyway), we're talking about the thousands of desktop boxes that a large corporation needs to administer on a daily basis.

    Further, when you consider the anecdotal evidence suggesting that a single, competent UNIX admin can handle more machines with less downtime than a competent Windows admin can handle Windows boxen--even without "set and forget"--the "TCO" starts to look favorable for UNIX.
  • Yeah, for people like me. When I drive a car with an automatic I can get started fine, but about the time to shift I automaticly push the clutch, and miss. By the time I remember I don't have to manually shift I've coasted to a stop, and the process starts all over a couple more times. (For a shift lever you just need a sympathetic girl who has her hand in the right spot, and let me move it to the next "gear" from time to time)

    Okay, I don't really coast to a stop, but I do attempt a shift and miss. And I have seen my dad reach down by habbit and move mom's hand. (Being a geek I'm not allowed to have a girl myself)

    If I have a choice I refuse to drive anything with an automatic, but sometimes I don't.

  • Tell that do one guy I know who twice put gas in his diesel van. (Accually his wife's van, which he rarely drove) Fortunatly he remembered at about half a tank both times, but he still had to tow it in and get the gas tank flushed for grabing the wrong pump at the gas station - the handle that normally is the right one on his other cars!

  • by bluGill ( 862 ) on Tuesday April 03, 2001 @09:30AM (#318044)

    Sure, you can save your company $50,000 by using linux of NT, but that is licensing costs. If you have to hire anouther admin because it is more work for your administrators to get their job done, then you haven't saved money.

    The cost to hire a compitent administrator are about the same for Linux, NT, other Unix, or anything else. The cost to train someone to do that job is about the same. It doesn't matter if you are using macs, windows, multics, Os/2, linux, solaris, or Os/390, the hard part of administrating the machine isn't learning the job it is knowing how to do it right.

    Linux doesn't save any money to a IT shop due to licensing fees becuase windows licesne fees are not that much. However linux can save a lot of money if the machines are set and forget. administrators are the expensive part.

    Money is not the most important thing to a bank, it is the ability to get customers their money all the time and accuratly refelect it on their statements. A bank wants to save on comptuer costs only if they are sure the ATMs are working. A phone company cares more about calls getting though then about the cost of comtpuers. I've seen several places with 2 Sun E10ks sitting right next to each other, one a hot standby incase something bad happens to the other. They have a 3rd E10 in a different city they can bring online quickly. They still wouldn't let me re-boot the master e10 despite having backups avaiable just in case. The data was far more important then considerations like computer costs. they told me that 1 minute of lost data is more then the cost of those machines.

    Linux isn't a factor because it is free (beer). It might be because it is free (speech). I have seen places with custom linux kernels when they needed it, but that is rare because that same shop had to manually merge in every security patch.

  • Using Microsoft products doesn't protect you from this- quite the contrary. It places you on an ugrade treadmill *forever*.

    Corporations were discouraged by Microsoft from upgrading from Windows 98 to Windows ME. But when Windows XP comes out RSN, Windows 98 will go on the dustbin of history and cease to be supported by Microsoft. Corporations will be forced to upgrade their entire networks immediately. I hope they're preparing.

    Brian
  • VA Linux is about as far from paying for software as you can get but they had one of the largest staff reductions of the year on a percentage basis. Software licenses are 1/100th of the cost of one employee. Remember that insurance often costs more than the salary. It may seem academically correct for companies to defer sofware costs to employee salaries but the fact is no-one is going to tranfer assets when they can liquidate the employees and keep even more cash. Computer scientists in industry are more like fuel than valuable contributions. Stay in academia.
  • Yes, I realize that this is a troll, but it serves to illustrate a point.

    Star Office is licensed under a "free beer" style license. You don't get access to the source code, but you can install it on as many computers as you like. Check it out [sun.com]. OpenOffice, Star Office's next generation, is available under the GPL.

    And don't think that Microsoft isn't worried about this. They know very well that once the customer's only reason to stick with a piece of software is "it will cost too much to migrate" that the software is essentially doomed. The short history of computing is littered with products that had tremendous market share and were eclipsed by less expensive (and often less able) competitors. Lotus 1-2-3, WordPerfect, and Novell Netware are probably the best known examples. Migrating away from these programs cost businesses a mint, but they still did it.

    Linux has a lot of advantages. Chief among them are that it allows developers and solutions providers to cheaply an inexpensively create turnkey solutions that are then deployable without license fees. So while it may not make financial sense for a company to chuck out existing systems in favor of Linux, it does make sense for new systems to be written using Linux and other freely available tools. By and large this is what is happening with Linux today. No one is replacing Windows with Linux across the board, but lots of shops are finding ways to implement Linux based solutions instead of Windows based ones.

  • This is a *GOOD* thing. Moving from Win95 to WinXP is as big of a move as moving from 3.1 to 95 was. It's a huge evolutionary step which is really going to boost the possible uses of our desktop computers.

    As far as the upgrade treadmill. Honestly, at some point the software does become good enough.

    I believe with XP we have reached this point.

    There has been a movement to upgrade to Win2k, but it is not as high of a priority as just completing day to day work. Also many companies at this point are just going to wait for XP on the desktop.

    Barring a complete overhaul of the way we use computers, I just don't see much motivating factors in jumping to the next version.

    There is also the same upgrade treadmill with Linux as there is with Windows, and the costs are nearly identical. The problem is *NOT* the cost of software licensing, but rather the cost of DEPLOYMENT.

    You would have had to be involved in a large scale upgrade to understand this. Imagine the number of people involved in planning, testing, rolling out, etc.

    Even with automated tools to push the installations, there is a signifigant amount of prep work involved.
  • What you say is partly true, but your reasoning is out of touch.

    The Recession occured between 90-92.

    The Migration from Netware to Windows NT didn't begin to occur until 96-97.

    But the cost cutting wasn't driven by the Bush Recession(especially considering that had occured years previous), but rather because of the new revolution occuring in desktop computing. The introduction of Netware 3.x, Windows 3.x, etc. had caused a huge increase in the signifigance, importance and usage of desktop computers in companies.

    Prior to this point they had generally just been somewhat distributed PC's running some DOS apps such as 123 and Wordperfect for personal use.

    Obviously as PC usage started to displace Mainframe usage in the companies, the costs assocaited with supporting these solutions went up. This was primarily because of all these individual solutions that grew up at the department level.

    In the mid 90's there was the movement to bring all these individual department solutions under the umbrella of corporate IT using the catch phrase "enterprise solution".

    This was also about the same time as Novell released their Netware 4.x product line. Which was notable for being one of the buggiest software releases ever.

    At this time Microsoft did introduce NT with a very competitive pricing structure as compared to Netware. But they did more than that, they also introduced a platform which could do more than Netware. They were quick to point out that you could also run your email server, database server, web server, etc. all on the same box.

    While you could sort of do this with Netware, the solutions were not elegant, and from a developer point of view a pain in the butt to create.

    So Microsoft won that war through a combination of competitive licensing, but also better product alignment. By offering the one size fits most all, they allowed for a reduction in support staff, which was part of what IT was looking for at the time.

    Today this new found integration of various solutions has resulted in incredible productivity improvements to companies.

    If you think Linux is going to come in and replace this existing solution, you had best think again.

    Right now we're not looking for ways to cut software licensing costs, because they are a small part of our budget, as is hardware procurement.

    Half our IT cost is custom software development, and thus are goal is to more rapidly create and deploy software.

    I do not see the tools out there today to allow us to do this well on Linux.

    Not to mention the tools to replace our existing productivity tools don't exist.

    If the RedHat salesman comes calling with numbers in hand, I'm going to ask him "Fine, but how will this be better than what we have now?"
  • "Further, when you consider the anecdotal evidence suggesting that a single, competent UNIX admin can handle more machines with less downtime than a competent Windows admin can handle Windows boxen"

    It is anecdotal.

    I've admined both Unix and NT servers. We have both Unix and NT here at our company.

    The problem with this comparison is that the two types of servers do different things.

    Is it less time consuming to admin a file/print server than a Database server? No
  • From a desktop admin side, most companies I have been at typically have one support person per 100 desktops. Then perhaps 1 server admin per 500 users.

    A lot of it depends on what services you are trying to offer.

    I was at one insurance company once with about 1500 employees and we had about 20 people supporting desktops, servers and the LAN hardware.

    At a bank I was at, the department I was in had 6 people supporting 80 users. But it was the stock trading floor. The remainder of the 2,000 employees in the company had perhaps 40 people supporting them.

    I don't know the size of the support staff where I'm at now, but I would imagine a ratio of 1:50 is about ballpark.
  • Ok, once again Malcontent is obviously showing his ignorance. Sigh.

    I didn't say the problem was Software distribution, that's actually quite easy as well under Windows using tools such as Tivoli, SMS, etc.

    I said the issue was Deployment. I then went on to talk about the various issues surrounding coordination, testing, etc.

  • Actually Tivoli is what we use, and it's worse than SMS. At least for software distribution.

    But again, you missed the point I was making. The costs have little to do with the actual file copy. That's the easy part.
  • As somebody's signature says "Linux is free if your time is worthless".
    __
  • Actually, each of these are pushing Linux to the front. It's just that the predicted time period was (and is) much too short.
  • Hmmmm... where I live the rate is closer to $30 Gs/year for Linux people.
  • Actually, you miss several facts. First of all, much free software actually comes from outside the U.S. where economic conditions are much worse. Secondly, you will have companies more willing to contribute than before. For example, my company is considering hiring someone to modify HtDig to do what we want it to, rather than paying for a commercial search engine. Commercial search engines cost about $75Gs a year, while we can hire a part-time programmer for $15/hour. It will take longer, but the savings will be tremendous. I can see this happening a lot.
  • Have you used the Gnome suite of stuff? Or StarOffice? Or any of the other office suites that are out there? If you know how to use Office, and can't figure out how to use Abiword, you need to get that lobotomy finished up....

    You will need to seriously train only a very small number of people (your current power users) if you were to switch to a new product. Most people will get by knowing how to open and save documents, and how to do basic editing. The rest will be able to survive on a 3-hour group class with handouts that tell you how to "use" the software. Not everyone needs to know how to cascade formulas across 4 worksheets, collect the results on a 5th, import that data to create a graph, then embed that graph in a document for circulation.

  • by TBone ( 5692 ) on Tuesday April 03, 2001 @09:36AM (#318067) Homepage

    1) Large companies don't always do layoffs to "reduce expenses". They do it to reduce the supply because of a forecasted reduced demand

    Then what do you call layoffs at a company that supplies nothing other than the manpower to operate the parent company? I call it cutting the bottom line.



    2) If we are talking about technical people being laid off, it will still happen: they don't know Linux and so have to be replaced

    There will still be computers that still have "I can't log on" or "I can't get my mail" problems, and we will still need helpdesk and cable monkeys to take care of them.



    3) MS licenses are not an operating expense, they are a capital expense (capital offense?). Meaning they already have money locked up in licenses

    Yeah, buyiong a piece of software is a capital expense. But owning the livcenses to a piece of software over the life of the company becomes an operating expense. If I run a company with 2,000 people in it, and use Office for my App Suite, then I have to stay at either the current or the previous version to maintain support from Microsoft. Older versions are not supported, and I can't afford to support an old piece of software on my own time. Dumping the software only loses money until the point where you would have had to relicense the new version to keep you supported.

  • by TBone ( 5692 ) on Tuesday April 03, 2001 @09:28AM (#318068) Homepage

    What company are you working for, then? The last two companies I worked for both spent well over a couple of million dollars a year in OS and Application licenses for the desktop. That doesn't include things like Oracle, Notes, or whatever that ran on the servers and weren't directly visible to the user.

    If I could save my company 2 million dollars a year, you better bet I'd be getting a raise and a promotion...

  • You are correct in your assessments!

    I think a lot of people forget that switching from anything to Linux involves substantial retraining costs, to say the least. And that retraining cost could run into the thousands of US dollars per employee.

    Linux is cheap, but training users is definitely not.
  • I think the whole issue of Linux acceptance outside of the server realm comes down to this: there is TOO much choice in how to set up Linux.

    Alas, this results in a major IT manager's nightmare, to say the least. Above the kernel level, the choice of various components you can install can result in non-standardization, which results in totally silly IT maintainance costs.

    I think this is why everyone has more or less gravitated towards to major commercial Linux distributions and IT managers are strictly enforcing a policy of which Linux components each individual workstation can have. That means you have to run either KDE or GNOME, not both.

    Hopefully, the Linux Standard Base (LSB) effort will bear fruit, which means there will be standardization in regards to every aspect of Linux. This should reduce system administration costs, to say the least.
  • It used to be "Nobody ever got fired for buying IBM" Then came the cheaper clones and M$ and "Nobody ever got fired for saving money" Now comes the even cheaper clones with Linux and its "Nobody ever got fired for lowering their TCO."
  • the costs incurred with those choices are incidental to the hardware, the development time, testing, deployment, help desk support, training, documentation, etc.

    This has been on my mind lately. I've noticed the costs of all of THOSE things seem to be a lot more on Windows-based machines than on the Linux machines in our small company. The hardware requirements seem higher to get the same performance, development seems quite easy on the Linux machines - at least as much so as it would be on the windows machines [web-based stuff which can also be accessed from other platforms, so long as a decent web browser is available on it], testing and deployment are simple enough, the Linux machines have thus far required a lot less help-desk support than the windows machines....you get the idea.

    On the other hand, I haven't yet tried Linux on an "ordinary user's" desk. I'm hoping to get a spare, older machine set up as an X terminal for all of the employees who don't have their own computers to get email, do web browsing, etc, with. We'll see how that goes...but I fully expect the support for this will still be a lot easier than another windows machine would be.

    The costs of one-time software license purchases ARE tiny in the grand scheme of things...but the cost of many purchased upgrades and extra support, hardware, etc. over the years does add up...


    ---
    "They have strategic air commands, nuclear submarines, and John Wayne. We have this"
  • Just out of curiosity, what's nsh? I recognize all the other tools, but not that...

    "If ignorance is bliss, may I never be happy.
  • by Guanix ( 16477 ) on Tuesday April 03, 2001 @09:17AM (#318088) Homepage
    Wait. With Microsoft, people pay for product and support. With Red Hat, they only pay for support.
  • What is an MSCE?

    --
  • I think the whole issue of Linux acceptance outside of the server realm comes down to this: there is TOO much choice in how to set up Linux.

    The Linux Terminal Server Project [ltsp.org] provides one solution to this. As far as IT issues go, it should provide a "win-win" -- you get to use standard, cheap PCs as a "thin client" AND you get the centralized control over everyone's configuration that IS loves so much (and that does help lower TCO).

    For a look at how this works in the "real world," read Major Law Firm Installation of Linux [linuxpr.com]. This provided a standardized KDE [kde.org] desktop for the administrative staff. Since the customer was a long-time WordPerfect user, the staff did not require application retraining, and only required minimal training on the Win* to KDE conversion.

    This also had the nice effect of changing an IT manager's nightmare into an IT manager's dream. The law firm ended up with a single point of control for all their desktops, which they could then even oursouce many operations back to Unique Systems [uniqsys.com], the company which did the rollout in the first place.

    What IT manager wouldn't like to be able to say to their boss "Look, for a small consulting charge and minimal retraining, I managed to cut our license fees AND support costs, preserve our legacy data and applications, outsource our administrative overhead, and I did this without purchasing any new hardware"?

    (And no, I don't work for Unique Systems [uniqsys.com], I'm just familiar with them from them from their good work with the Toledo Area Linux Users Group [talug.org], and from considering just this setup for a former employer.)

  • by rw2 ( 17419 ) on Tuesday April 03, 2001 @09:16AM (#318091) Homepage
    His column [infoworld.com]

    --

  • > Except with the smallest of companies, software is rarely a major expense for companies, especially OS's.

    On paper, at any rate. In practice, everywhere I've ever worked was really tight-assed about spending pocket change during tough times, presumably because everyone was afraid of getting a chewing from above every time they dropped a dime.

    Also, correctly or no, employers tend to think labor is "free" in the sense that since you're already on the payroll they already have to pay you anyway. Most employeers ask their employees to do 4-5 times as much as they can realistically do, so adding on "learn a new trade" or "manage a bank of new machines" for your spare time is not likely to faze any boss I ever worked for.

    Your argument has rational force, but PHBs and PHMMs are about as irrational as it comes.

    > It may be cheaper for a company to pay a few grand for Windows, then save tens of thousands by hiring a generic Windows admin.

    Or a tribe of them?

    Reminds me of the joke about the town in the wild west that telegraphed Washington for help during a riot. The citizens were dismayed when a single US Marshal stepped off the train. His response was, "You've only got one riot."

    --
  • > In every company I have ever worked for, the attitude has always been to spend what you need to in order for your people to be productive.

    Whoa! I need to consult you about job hunting tips. I've never met a boss who wouldn't waste a month having high-priced engineers dig a ditch by hand because he was too cheapskate to rent a backhoe.

    --
  • My Uncle's company is growing and they looked at the licence compliance issue with MS Office. For each 10 people they would need to shell out $4,000. Or they could just give people a copy of StarOffice and by a new Hazmat suit (they do clean up operations). So the office is going over to StarOffice. Not too shabby.

    --Peter
  • All I see is a shift from products to services. Instead of creating a total package which is 99% complete and charging $1000 a head, you download an 80% complete package for free and pay someone (internal or external) to add the other 19%.

    The 1% of course is that feature which never gets done :)
  • Any bets on how long it will be until we're back at "Microsoft's inability to deliver" again?

    I don't know. What are the latest guesses on when .NET and WindowsXP will be out? :)
  • Same thing we always did...smile politely and keep them away from the servers.

  • Microsoft's inability to deliver was going to push Linux to the front.

    Then the dot com revolution was going to do it.

    Then the consumer bandwidth revolution was going to do it.

    Then the globalization of computing was going to do it.

    Now an economic recession is going to do it.

    Any bets on how long it will be until we're back at "Microsoft's inability to deliver" again?
  • With MS Turning Screws on Customers the way they are, I would imagine that many mid size company's would turn to free software, instead of being faced with up to $10,000 per unlicensed product.

    It needs to be remembered that the cost of a, per machine, licence is not just how much you pay for it. But also the cost of an administration system to ensure that licencing is kept up to data and that proof of licencing can be produced.
    Very few payware software companies offer a true site (or organisation) licence.
  • For a large company the actual cost of purchasing the software may not be the most expensive part of switching. Things like retraining, time spent rolling out the software, technicial support, document convesion can add up to alot of money.

    Following the "buy Microsoft" route can mean doing this every 18-24 months. With the added disadvantage the the methods used last time to ease rolling out a package across the network may no longer work.

    Most IT people do not know very much about writing a business case. Most management types don't know technical speak. As a result the tech who thinks it is obviously cheaper and better to use free software has a hard time convincing management.

    Also the tech has a hard time getting mangement to understand that just cos it's easy to install something from a CD on a standalone machine dosn't mean that it will be easy to install the same program on a network.
    Especially in the worst case senario of requiring manual configuration after install.
  • Corporations were discouraged by Microsoft from upgrading from Windows 98 to Windows ME.

    Probably because all of the "improvments" between 98 and ME are aimed at the home user and offer little to the corporate admin.
  • This simple premise makes upgrading and maintaining remote windows machines awfully difficult especially considering the fact that a typical windows app spews files in all kinds of unpredictable directories and the fact that registries have to be kept in sync.

    Another thing you tend to find is application programmers asuming they can write to any file with programs sulking if they can't open a file they only ever need to read as R/W.
    I would place the blame more on app writers than Windows. Since it is possible to write well behaved and structured apps, even under Windows. e.g. most of the app goes into one place (which could be shared network area), any changes to Windows system area or registry are well documented and sensible defaults are chosen if the program finds it is lacking registry keys. However most Windows programmers prefer to produce "sphagetti"...

    Overall tools such as rsync, httpsync, cvsup, cfengine, nsh, ssh, perl etc are so great that a single sysadmin can manage a widespread network if trained well enough.

    Assuming that "mount server:/apps /apps" won't do fine anyway :)
  • Yeah, admins are expensive, but at thousands of dollars for a 5 user license pack for Win 2000, the difference starts to diminish -- remember, we're not only talking about the big iron in the back room (which, frankly, is likely to be running a *NIX anyway),

    You can have Windows "big iron", e.g. the way Microsoft were running their DNS servers...

    we're talking about the thousands of desktop boxes that a large corporation needs to administer on a daily basis.

    However we are talking rather different admin models here, e.g. a unix workstation is not going to expect the end user to be performing sysadmin tasks, rather than getting on with their real job.
  • The talent pool for MS software is broader than that of Linux

    With some people producting MS software "talent" is not really the best description. It might help if they understood the difference between software and pasta!
  • On the other hand, I was hired at a place that is (mostly) MS driven, but only because I knew Linux. Any Linux person can learn the MS software in a reasonable amount of time.

    Though they might find some things a lot of hassle. e.g. having to manually create home directory shares on NT.

    On the other hand, teaching a windows person how to do anything with unix is not as easy.

    Could even be a case of replace "Linux" with "most, non Windows operating systems".
  • But the question, really, though is what the environment is like. I sure as HELL would never, EVER, not in a million YEARS put anything but Windows on one of the computers here at the university I work at.

    Just as "time flys when you are enjoying yourself" it drags along when you are having a bad time. If just under a decade feels like 100,000 times as long then the previous poster must be having an utterly awful time :)
  • the constant reinstalling of the OS because the user installed some simple software and trashed the system, the "support" or lack there of from Microsoft (Oh, you have istalled software from [insert non-microsoft company here] and that ruined your computer, so bugger off and call their support)

    It's not unknown for this to happen even with Microsoft apps. My favourate was if you installed Microsoft Office on a Windows 95 machine then Spot scanners wouldn't work.
  • The last two companies I worked for both spent well over a couple of million dollars a year in OS and Application licenses for the desktop.

    How much did you also have to spend in man hours administering, recording and filing those licences too? Would have been far less hassle to have a "site licence" which simply said you can use software X in your organisation. You could consider "free software" as being a "site licence" which costs nothing and also allows people to use the software at home.
  • You put Gnome in front of the average user, and they're lost. I hate to say it, but the average user gets confused with the 'Save' command.

    Yet the same user who's only ever driven a car can probably manage to drive a small truck without much fuss.
    If they get this lost how are they going to cope with the next version of Windows which has the UI changed?
  • Switching software platforms means you need to either hire new specialists or retrain your techies... either costs money.

    It's kind of tricky to not have constant switches in software platforms where you use a platform which changes ever 2 years or so :)
  • Games: You want to play all the newest games on a PC (which is a unique market compared to the console market), you have one choice. Windows.

    What has this to do with corporate networks?
  • Try CUPS with GTKlp. I'll take that over windows printing anyday thank you very much. I mean, send duplex print jobs, specify print trays, even 4 pages to 1 sheet of paper. And no, the windows print drivers do not support those options.

    Also Windows printing is a poor standard in the first place. e.g. no ability to limit paper size choices. So people set up documents for the wrong size paper and either have them auto deleted or jam up the printer.
  • This is the whole point!! Do you realize how completely clueless 99.99% of all users out there are? They can't use the clipboard in a word processor and it takes them 30 seconds to find a scrollbar. Do you think they want to type "rpm -Uvh" in some shell?

    No you are missing the whole point. Installing software is a job for the sysadmin. NOT the end user!
    If someone has a company car should it be servicable using the driving controls because the driver dosn't know how to hold a spanner? Do you force a mechanic to service the car from the driving seat because some driver might sometime think that being able to drive a car qualifies them to be a motor mechanic?
    No you don't because this is self evidently daft. Indeed most of the problems with Windows are probably down to encouraging end users to fiddle with things they know little about, rather than getting on with using the computer as a tool to help in their job.
    The end user dosn't need to know what "rpm -Uvh" or "USER.DAT" means any more than they need to know what size bolts hold their car engine in place.
  • End users shouldn't install software in a corporate setting. End users tend to install melissa.

    Or manage to break things, or install software without making sure that they have a licence and the paper trail to prove they have a licence.

    However, the people doing upgrades might like the fact they can upgrade every machine with software package foo by running a script that will automatically install it on all the machines rather than clicking n times per machine.

    Also without needing to physically go to the machine, not much of a problem if it's 2m away, an annoyance if it's 1km away and a very big problem if its 1,000km away. Not only that the install can be done without needing to bother the person actually using the machine.
    There is a lot of difference from a home machine which is end user administered and even the simplist of networks. The same way that there is a difference between someone tinkering with their own car and running a corporate vehicle fleet.
  • I'm sure that all of you who have done tech support will identify with this. It's not that people wouldn't learn if they tried, they just won't try unless they have their hand held every step of the way.

    It's actually a wetware/peer preasure issue. If someone could only operate one type of car the'd be looked on as a fool. (Even though there are a few silly things done with cars.) But it can be almost a kind of perverse pride for people to be only able to operate one kind of computer.
  • The radio isn't a vital system. Everybody places the gears in PRNDL order. Unless GM wants to get sued by a lot of people, they must place the pedals clutch, break and gas from left to right in that order.

    An automatic car with a clutch pedal, interesting concept. Not unknown for automatic cars to have an auxillary brake where the clutch should be though.
  • Actually, back in the days when cars came standard with this wonderful thing called a manual transmition, the gear order could have a variety of permutations, most commonly altering where the reverse gear would be.

    In many parts of the world automatic transmission is uncommon. There is also semi-automatic, has a control similar to manual but no clutch pedal.
    Manual gearboxes also vary between 3 and 5 forward gears.
  • Consider this, and this. Businesses not only must wade through enormously complex licensing from Microsoft, but they run the very real risk of being audited - with the price of running an unlicensed copy at $150,000 per instance ! Massive effort is required to maintain software licensing compliance, with no real guarantee that the auditors couldn't find something.

    How much does this cost, in terms of man hours? How many sysadmins would the same money buy?
  • For example, my company is considering hiring someone to modify HtDig to do what we want it to, rather than paying for a commercial search engine. Commercial search engines cost about $75Gs a year, while we can hire a part-time programmer for $15/hour. It will take longer, but the savings will be tremendous.

    Does the $75G figure cover having it customised to do what you want it to do?
  • It seems all good, free software to cut down in licensing fees, but what about the cost to retrain staff?

    You are still going to have to retrain even if you stick with Microsoft. Since every couple of years or so Microsoft will be bringing out a new version.
  • Windows XP is scheduled for the autumn I think, and probably won't slip as much as Win2K since it's not as radical an upgrade as Win2K was. .NET, well, that one is definitely tougher. The first beta of VisualStudio.NET is out, but it's definitely a *BETA*. My guess is at least 12 more months until anything substantial happens on that front.

    My guess is also that in 12 months, nothing *substantial* will have happened in the "market shares" of open source vs. closed source.
  • Point 1 about HTML editors.. tsetem replies "I prefer VIM anyday". Point 4, installations, tsetem replies "rpm -Uvh or rpm -Fvh".

    Hello? This is the whole point!! Do you realize how completely clueless 99.99% of all users out there are? They can't use the clipboard in a word processor and it takes them 30 seconds to find a scrollbar. Do you think they want to type "rpm -Uvh" in some shell? That's the whole point that the original poster was making - that it's not just the COST of software that matters.

    Do you have any idea what the cost of getting these 99.99% of clueless users to the skill level where they could do anything productive with Linux - at it's current state? License costs are totally meaningless in that perspective. Mac OS X is pretty close to what a USABLE UNIX should be. I'm sure most users could switch to that without any big learning curve. But Gnome and tcsh on Linux, not a chance!
  • Wow! According to your definition Jesus was the biggest communist in the world.
  • I don't think he was talking about large corporations (they are too conservative), I think he meant small to medium sized companies who can not afford the lawyers to fight MS.
  • Software distibution is much harder to with windows then unix. This is mainly due to two factors. One unix has been doing it for twenty years now and it has a wealth of proven and tested tools. Two is that primary assumption of windows is that there is only one user and that person owns the computer. The primary assumption of unix is that there are many users and root owns the computer. This simple premise makes upgrading and maintaining remote windows machines awfully difficult especially considering the fact that a typical windows app spews files in all kinds of unpredictable directories and the fact that registries have to be kept in sync.

    Overall tools such as rsync, httpsync, cvsup, cfengine, nsh, ssh, perl etc are so great that a single sysadmin can manage a widespread network if trained well enough.

    As far as retraining goes well every major upgrade of windows is sufficiently different enough to retrain your employees anyway. Besides if you can port your app then you are 90% there anyways.
  • I think kylix is pretty damned hot. It's way more productive then VB. As a bonus you can run your apps through X terms.

    If MS is not lying about .NET (yea right!) then you should have nothing to worry about. .NET will allow you to build apps in anly naguage which will run on both windows and linux. If MS is lying about that then you have tools like perl or python which are very productive and wxwindows which is very cross platform.

    It may be true that you have a staff full of VB developers but remember you are going to have to teach them VB.NET. VB.NET is so different then VB that they may as well learn another language like object pascal (which ironically will seem more natural to them then .NET).
  • I collect posts like this and show them to high school students who are asking me what to learn. I tell them "see linux programmers and sysadmins get more money then windows programmers and admins" I have whole bunch of these types of statements from the top brass at MS. "See even Bill Gates admits linux professionals get paid more".

    Please post more things like this. Next time don't post anonymously and provide a link it's much better for me that way.

    Thank you so much!
  • I love posts like this!.

    I save them and show them to high school students who are deciding what to learn. "Linux professionals get better pay then Ms professionals" It makes great advertising.

    Thank you very much.
  • If your employees are that dumb (I don't doubt it) then you'll have to retrain them for XP anyways. In fact once the .NET hits you'll have to reatrain your entire company on how to use rented software. Most companies I know regulary send their employees to training sessions. I really don't see why they couln't adopt to openoffice or Koffice or whatnot.

    Really if your employees are that stupid get macs for them. Save yourself millions just on helpdesk calls.
  • WOW you never actually tried to get support from MS did you?
  • network shell it's a commercial product and an excellent one at that.
  • Unfortunately I have intimate knowledge of software distribution including the horrendous and expensive monster that is called SMS. Maybe one day MS will produce a version of SMS that works as advertised but they seem to realize that they just can't do it and are moving to terminal services and of course the panacea known as .NET which will magically solve all of your software problems.

    Tivoli is actually not godawful but it sure is godawful expensive.

  • You don't have to you could script the whole thing. Many of the tools like cfengine support all different flavors of unix so it's really not that bad of a problem. Certainly much easier then trying to manage a network of win95, win98, nt3.5, nt4.0, ME, w2K machines. I would say it's easier to manage a network of mixed linux, solaris, and freebsd machines is easer then trying to manage a network of mixed windows versions.
  • Wal-Mart had some of its strongest growth and expansion during the recession of the early 90s. While the economy and the retail sector in general was hit hard, Wal-Mart steamrolled on.

    As a company with the price advantage in a completely different industry, we are also finding a lot of new business lately.
  • Everybody places the gears in PRNDL order. Unless GM wants to get sued by a lot of people, they must place the pedals clutch, break and gas from left to right in that order.

    Well, the rest of the world knows that Merkins are stupid, and can't learn new things. In the rest of the world, where most cars do not have automatic transmissions, some have reverse on the left of the gate, some on the right. Some have a lift collar to select reverse, some have a push down to select reverse, some have no safety lock mechanism. Most have the gear shift on the floor, some have it on the dashboard, some have it on the steering column. Some have four forward gears, some five, some six; most 4x4s have at least ten, usually with an additional manually operated transfer box. Any non-American driver can get into any of these vehicles and drive it without retraining.

    So I think, really, your example does not work - except, perhaps, for Merkins.

  • by superid ( 46543 ) on Tuesday April 03, 2001 @09:36AM (#318146) Homepage
    A few weeks ago there was an episode of The Lone Gunmen that I really enjoyed. They were searching for the magical car engine that ran on water. The Evil Nemesis (tm) was of course from the oil company. During part of the climactic speech, he said that basically "great, a car that runs on water....but you *still* need oil for lubricating, making plastics, making roads Hahahahahaha!(evil laughter)" and I immediately drew the parallel to the software industry.

    I'm basically a web developer now, 3 tier, database, middleware, webserver. Any or all of those tools could be free or $$$ depending on which group of developers in my office you talk to (ie. some want MySQL, some want Sybase....Apache vs IIS, etc) but the costs incurred with those choices are incidental to the hardware, the development time, testing, deployment, help desk support, training, documentation, etc. I'll bet if we were 100% LAMPS, it would save us less than 10% (but I'll still try :))


    SuperID
    Free Database Hosting [freesql.org]

  • My company constantly bitches about CVS. We are about to move to Perforce and they are putting together a list of requirements that it will have to meet. So basically I ask: Why not take the requirements and fix cvs? Nah, we'd rather pay some company for a product we cant fix and then bitch when it doesn't meet our requirements.
  • I see in a few months people will start to admit that perforce cant do what they want either and then we'll end up with some other crap, rather than just making what we want, submitting the changes back to the cvs tree and letting them fix the damn bugs.
  • Well, so the old story goes, if they really are decent *n*x talent, one warm body will go a lot further than it would in a Windows shop.

    Anecdotal evidence: I work in a software development shop. Mostly Linux, some OpenBSD, some Solaris, two or three Windows laptops, one Macintoy for the interface designer. Pound for pound I spend a lot more time fixing the Windows boxen; next comes Solaris; then the OBSD and Linux machines which require next to no day-to-day fixing. I do all I can to ignore the Mac.

  • Possibly because the stock market as a whole has taken a real dive over the last 2-3 weeks. The Dow dropped over 300 points before lunch today and the NASDAQ ain't been looking to hot lately either. Somehow, I suspect that the root of all this isn't that every company in the US is going to go bankrupt this year.
    _____________
  • My point was that the differences between MS Word, Word Perfect, KWord and StarOffice are largely cosmetic. Each of these applications has a tool bar that contains pictures representing the most common operations, a large white space in which to write, a menu bar that contains a file menu for opening and saving documents and so forth. However, training users to recognise that the same basic concepts are to be applied would take far longer than a car dealer spends explaining the features of a new model. I consider these basic operations to be equivalent to the clutch, brake, gas, steering whell etc. The differences between them are more like turning on the high beams. Some cars you pull the turn signal (Chevy), some cars you push the turn signal (Ford), some older cars you tap a button with your left foot.

    There is actually a sizeable portion of the population who would be very confused if you were to replace the word Start with a picture of a foot or a gear. Even though the reason to click the lower left corner remains to log off. In reality, from a user's perspective a bash or command.com CLI are strikingly similar. You type the name of the program you want to run. Save files in directories and move between them with the cd command. Likewise, the window environments are all quite similar. It would seem to me that it would be much easier to move from Windows to Gnome or KDE than to MacOS, but people seem to think that MacOS is easy to use so that's not given any consideration.
    _____________

  • by lizrd ( 69275 ) <adam.bump@us> on Tuesday April 03, 2001 @11:06AM (#318155) Homepage
    If automobiles were this way, you'd have to completely relearn how to drive on a completely different road system if you wanted to switch from Ford to Chevy.

    If people reacted to automobiles the way that people react to computers they would need to be taught how to drive again when switching from a Ford to a Chevy. I can just imagine it.

    "Ok, now. This car is very similar to the one you had before. Remember, the wiper lever is now on the left side of the steering wheel. It is not on the right like on your Ford."

    "In this vehicle you will have to locate the air conditioning controls which are above the radio. They are no longer beneth the radio. No, I said above the radio. No, that's the ash tray. Move your hand up. Further up. up. up. up. There. I know, this car has knobs instead of sliders. Yes, that is the heater control. I know it looks different. I promise you, it is the heater control. You can do this. It's not that hard to learn."

    I'm sure that all of you who have done tech support will identify with this. It's not that people wouldn't learn if they tried, they just won't try unless they have their hand held every step of the way.
    _____________
  • I'm usually a Linux/Free software advocate, however...

    I see some major flaws in the logic here. If my company were to change over to a new platform, we would probably spend millions of dollars changing over the rest of our software, that it would negate the money we would be saving with the free OS. Unfortunately, we are sooooooo MS dependant, especially here in engineering.

    I can see change overs happening slowly, starting with the larger, more mission critical, more expensive applications, and eventually working down to the individual user's desktop. But that will take years.
    • If the RedHat salesman comes calling with numbers in hand, they will listen.

    The RedHat salesman? The guy who comes into Megacorp and tells them that they only have to buy one $49.95 copy and install it on every one of their 50,000 servers. Right.

    For good or bad, this Free Software is a different business model. There certainly won't be much room for software sales types and other middlemen.

    If Megacorp is to be sold on Linux, it won't be salesmen who do it. Maybe a RedHat salesamn will come through with an attractive Enterprise Support and Training deal, I dunno.

    What might do it is when people install Linux behind the back of the IT departments, like you say happened with Novell. If people start to get lots of work done with File and Print services based on Samba with excessed hardware lying around, then maybe those CIOs will notice that they really don't need to buy licenses for this anymore.



    ---

    • The IT scrubs can get Linux into production here and there in the under-the-nose fashion, but nobody is going to stop paying MS site licence fees until someone on top makes the big decision. This can happen if there's sufficent pressure to reduce costs.

    You can pretty much forget about CIOs, for the most part, being convinced that Linux will save them lots of money. There just isn't the money floating around to drive the schmoozing and the magazine articles that convinces them of anything.

    What'll happen is this. There'll be one of those brain-dead-we-can't-trust-middle-management-to-spe nd-any-money-at-all-because-they-are-all-a-bunch-o f-idiots, edicts like "All IT hardware procurement is frozen. No purchases except for replacement of critical systems is approved until further notice or the signature of 2 executive Vice Presidents. No exceptions." Someone will notice that John's department is deploying good infrastructure still with excessed hardware and Linux while Jerry's department is just whining all the time. John gets put in charge of Jerry's department and the amount of Linux in the organization doubles and a bandwagon starts which the CIO jumps on board.

    It won't come from the top down, that's for sure.



    ---

  • But I bet you the lion's share of profits of Mobil and company come not from lubrication, but from the fuels that they provide.

    Speaking of Mobil, the McLaren F1 team uses Mobil unleaded. If it's good enough for McLaren, it's good enough for me! (;
    --
  • by mr ( 88570 ) on Tuesday April 03, 2001 @12:25PM (#318166)
    I guess only time will tell if the use of Free Software is as much an economic advantage as people have been making it out to be.

    Free Software has a cost. That cost is the risk of your IP if you are not careful.

    Look at how Corel 1 [slashdot.org] and 2 [slashdot.org] had trouble 'getting it right' The risk to Corel is great enough vs the profit potentional, they are trying to sell off the Linux version they make. Free Software is not worth it.

    Open Source software (something the most visable supporter of Free Software RMS [stallman.org] wants to not be associated with. The Open Source Movement -- I do not support it. [mercurycenter.com]) *CAN* protect your IP. Understanding HOW you can have protection for your hard work AND work with Open Source software is the key to protecting your IP.

    Free Software wants to destroy IP. IP is the backbone of the information economy. Free Software does *NOT* have an economic advantage when it comes to IP. Open Source software *CAN* provide IP protection and therefore economic advantage.
  • by rodentia ( 102779 ) on Tuesday April 03, 2001 @09:35AM (#318173)
    I've noticed a surge in interest in the open-source foundations of my projects at work. We'll be presenting on Linux, Apache, Tomcat and Cocoon at an upcoming installment of the CTO's roundtable. There is considerable curiousity about how I'm funding my work in the current budget environment. The answer, of course, is that the stuff is free and runs on hand-me-down hardware.

    <herb>It looks here as though your only costs are salary....</herb>
  • by psin psycle ( 118560 ) <psinpsycle.yahoo@com> on Tuesday April 03, 2001 @09:27AM (#318178) Homepage
    I think one thing that is slowing the adoption of "free software" is that it is very difficult to get hard numbers on the actual cost of switching from one software package to another.

    For a large company the actual cost of purchasing the software may not be the most expensive part of switching. Things like retraining, time spent rolling out the software, technicial support, document convesion can add up to alot of money.

    Most IT people do not know very much about writing a business case. Most management types don't know technical speak. As a result the tech who thinks it is obviously cheaper and better to use free software has a hard time convincing management.

    One of the good things that Microsoft provides is business cases. If you want to move to a MS product, all the informaiton that you need is made available. (if it actually works as stated is another argument) On their website you can often find migration guides and business cases.

    This is an area where free software is greatly lacking. We need to build business cases to prove our case.

    I am currently taking a course that trys to bridge the gap between the tech speak and the manager speak. For my term paper for Emerging Technologies I am going to be building a business case for the adoption of OpenOffice in the enterprise. (www.openoffice.org) I will gladly donate the business case to Sun to use to help promote OpenOffice.

    I would encourage anyone who has to do a term paper like this to pick a free/open source solution to try to support.

  • by Golias ( 176380 ) on Tuesday April 03, 2001 @09:38AM (#318219)
    Also

    4) 5000 people at $40k each is $200,000 per year (plus you save tens of thousands in seat licenses for your software, because there are fewer people using it.

    5) Switching software platforms means you need to either hire new specialists or retrain your techies... either costs money.

    Bottom line is that a company will change to open source for one reason and one reason only: if they are convinced that they will make more money using it. In every company I have ever worked for, the attitude has always been to spend what you need to in order for your people to be productive. Whether that means dumping huge cash into an Oracle database, or hiring a Postgres DBA, productivity matters more than expenses.

    A CEO (or any manager, for that matter) who does a really good job of cutting expenses (mostly via layoffs) gets a reputation for being a great "hatchet man", and all of his job offers start coming from sick companies that need to make cuts.

    A CEO who does a really good job of raising productivity and sales gets a reputation that lands him powerful jobs in healthy companies. Think about it: If you were a corporate executive, which type of CEO would you prefer to strive to be?

  • by hex1848 ( 182881 ) on Tuesday April 03, 2001 @09:22AM (#318224) Homepage
    With MS Turning Screws on Customers [slashdot.org] the way they are, I would imagine that many mid size company's would turn to free software, instead of being faced with up to $10,000 per unlicensed product.
  • by SubtleNuance ( 184325 ) on Tuesday April 03, 2001 @04:28PM (#318229) Journal
    It's rather ironic that business' knee-jerk reaction to situations like this hurts their chances of making more money in the long run.

    Its not ironic - its criminal. Were talking about real people here, people with lives, families and futures. The reason were seeing this reaction to this minor 'slow down' (whatever the hell) is because Corps are no long pretending to have any priorities other than profit. Quarterly Profit as a matter of fact. If it looks like profits may (*shudder*) decrease - then the pricks go into 'crisis mode' and start layoffs. Because the Deep Pockets on wallstreet need Big Profit Now(TM) - and they dont care otherwise... if they loose confidence in XYZ stock because this quarter is bad - they sell - if they sell, then the CEO and the top 400 people in a Fortune 500 company loose big dollars on their stock options... so they are VERY motivated to make every quarter better than the next ... "the future" be damned, these top '400' people will be rich beyond imagination before that ever matters... the whole scheme is ready to eat itself... current Capitalist Economics are about to reveal the failure in their basic logic... this whole episode is inevitable. Unfortunately the rest of us are along for the ridiculous ride

    So, whats the Solution? Read .sig

  • by yoink! ( 196362 ) on Tuesday April 03, 2001 @09:38AM (#318233) Homepage Journal
    Today's Linux Desktops are getting more and more user friendly all the time. KDE2 and Gnome have the equivalent of a start button, and with many office suites in development there are a plethora of options for all those MS-Word mongers who happen to need Office Premium edition with the free spa day.

    Seriously though, there will always be a transition period, we can't really help that, but it's worth it. I'm not just advocating Linux here, but all alternatives, Linux just happens to be a nice example. Sure, we'll see a learning curve, a learning curve softened by the millions of dollars people save with free software. There are so many advantages to using other OSs and apps in workplaces I can't even begin to list them. People are scared of change, I'm scared of change. It took me two years to move to Linux; ie. to feel comfortable that I didn't need anything else. It's to be expected. But just because we're going to need to relearn a few things doesn't mean we can't start right now!


    yoink
  • by Auckerman ( 223266 ) on Tuesday April 03, 2001 @09:47AM (#318248)
    I often hear about the fact that "free" software costs less to buy with hardware as a reason for switching to Open Source. This is too narrow a view to put forth. Cost is not an issue, its what you can do with it. There are three things major reasons you buy a computer for: Games, Server, Workstation.

    Games: You want to play all the newest games on a PC (which is a unique market compared to the console market), you have one choice. Windows. Yes I know, MacOS has improved dramtically to the point that it is viable and Linux isn't that far behind, but face it game performance is usually better on a PC and there are more high end graphics orientated games for WIndows than MacOS and Linux combined.

    Server: Here is where Open Source has a large following. Apache, Samba, mySQL, Perl, and other such OSS technologies make Bill Gates wake up in the middle of the night sweating, and for good reason too. But, there are some things they don't do as well as something like Win2K does, but in general are more flexable and certainly significantly cheaper than the alternatives.

    Workstation: This is the majority market for computers and the biggest expense most businesses (and people) are going to make. Here Linux lacks some key ideas.

    1. No really solid HTML editors

    2. Poor application linking

    3. Poor printing services

    4. Its harder to update anything on Linux than Windows and MacOS

    5. Poor graphics support

    6. No unified GUI (KDE, Gnome, who cares, just make ONE of them work)

    7. Each Linux variant ships with security holes

    I could go on and on, but you get the point. Not only that, Linux has to deal with misconceptions in the workplace (its ONLY a server, its a hobbiest OS, If its free it can't be any good, It doesnt have any good Office compatible software...yadda yadda)

    People want something that they KNOW for a fact will work. Windows may have its cavets (instability, high cost of maintance, et al) but at least it runs Word, it connects to a custom database, and company can call MS (or the OEM) and get help when it dies. In a word, its "simpler" to just choose something convient than it is to built a custom system from components and then deploy it to every desktop in your business.

    If you want Linux to suceed, make a distributation that is NOT a server (meaning fix whats written above) that has a significantly better UI and easy to make custom install CDs (and I mean EASY), then you will see LInux being used.

  • by NineNine ( 235196 ) on Tuesday April 03, 2001 @09:21AM (#318260)
    Except with the smallest of companies, software is rarely a major expense for companies, especially OS's. Usually, a MUCH higher cost is associated with development and support (paying salaries, contractors, etc.). Whether the TCO of Linux is really lower than that of proprietary software is still up in the air, I think. It may be cheaper for a company to pay a few grand for Windows, then save tens of thousands by hiring a generic Windows admin.

  • by Petrophile ( 253809 ) on Tuesday April 03, 2001 @10:12AM (#318268) Homepage
    At one time the corporate PC LAN was ruled by Novell NetWare. At it's peak in the early 90s, it had about 80% of the "LAN Server" marketshare. It wasn't the most stable or capable product, but it provided a very managable system for file and print needs, and was often expanded for departmental e-mail service or database needs. Often these Novell systems where installed under the nose of the main/mini-centric IT department in a guerilla manner by departments.

    It was also pretty expensive by modern standards, costing something like $1000/seat/year.

    Meanwhile Microsoft had been slogging along with about a 10% marketshare for a OS/2-based product called LANManager. Even though LANMan had some nice features like TCP/IP support, Microsoft literally couldn't even give away (and they tried).

    Then, the early 90's recession hit, affecting certain corporate headquarters-heavy areas like California and the East Coast especially bad. For those of you who are relatively new to the job market, here's what during a recession: The corporation cuts costs to please Wall Street. An prime example of a "cost" here is the IT department and IT projects.

    So, as new generation of IT managers took hold, what they found was a mess: dozens of different mail systems and departmentally managed networks. Different warring tech support groups backing different technology. Hundreds of little Novell servers out in closets managed by relative amateurs.

    A few things happend as a result: 1) PC and PC network management was taken from the departments and centralized under the main IT department, 2) IT support was outsourced (it was virtually impossible to get a perm sysadmin job back then) and 3) Aggressive measures were taken to reduce licence costs.

    Meanwhile, the Microsoft salesman was standing at the door with their relatively new Windows NT product that carried a very attractive price. If you were large enough, they would negotiate a licence costs that was less than 1/3 of what Novell wanted. No matter that the file and print was less managable and less scalable, you could also use it as an application server to centralize (say) e-mail and/or move it off of those expensive mini's and unix systems.

    The result was devestating to Novell's business. Even though Microsoft's prices gradually went up, and Novell's went down, Novell was left with something like a 20% marketshare when it was all over.

    So, now we have a situation where Microsoft's prices are going up even higher and IT spending is due again for another round of cutbacks and centralization. Well, those licence costs might have seemed relatively inconsequential when the budget was expanding and you had things like Y2K and webification to worry about. But in a static or declining budget, they will stand out. And CIO's who are paid bonuses to cut costs will notice. If the RedHat salesman comes calling with numbers in hand, they will listen.
  • by CyberDawg ( 318613 ) on Tuesday April 03, 2001 @09:18AM (#318289) Homepage
    It's great to hear that some of the Linux-related companies are not suffering from that dot-com disease that says, "profits don't matter -- only market share and total sales." It's like the Japanese semiconductor companies in the 1980's ("we may lose a nickel on each chip, but we'll make it up in volume"). Unfortunately, as these companies crash and burn, they're hurting other businesses in the market, too. Take servers and routers. If you're looking to expand your business right now, are you going to buy new equipment from VA Linux and Cisco when you can get close-to-new stuff from bankruptcy auctions at dot-coms for pennies on the dollar? Nope. It's going to take a while for the ripples to die out from all of the dead companies run by people who didn't understand the basic tenets of profit and loss.
  • by BillyGoatThree ( 324006 ) on Tuesday April 03, 2001 @09:23AM (#318294)
    "Think about it: with companies laying people off and cheapening up, whats better than free software? They could lay off 5,000 people, OR quit shelling out that much in MS licenses and pay salaries!"

    1) Large companies don't always do layoffs to "reduce expenses". They do it to reduce the supply because of a forecasted reduced demand.

    2) If we are talking about technical people being laid off, it will still happen: they don't know Linux and so have to be replaced. Of course, firing an MSCE for being an MSCE might appeal to some...

    3) MS licenses are not an operating expense, they are a capital expense (capital offense?). Meaning they already have money locked up in licenses--dumping the software loses that money. I know, I know--it's the fallacy of sunk costs. Nonetheless, some business people work that way.
    --
  • by glenkim ( 412499 ) on Tuesday April 03, 2001 @09:16AM (#318299) Homepage
    Pretty cool thought, but it still seems like it might be a fluke. Was the recession taking place during the time span reported in the earnings reports?

The fancy is indeed no other than a mode of memory emancipated from the order of space and time. -- Samuel Taylor Coleridge

Working...