In my last post we talked about comparing Linux and Windows on the basis of bugs and/or vulnerabilities. Not surprisingly this generated lots of discussion. One of the comments that caught my attention was a one-liner that said "No matter. 2005 is the year of Linux. We have Microsoft right where we want them."
Well, in my opinion, I don't think that this is the case. However, what may be a more interesting discussion is whether 2005 is the year of commodity hardware. From the discussions that I have had with IT Managers in the past, there seems to be a recurring theme, they want to reduce the amount of money that they are spending to run their IT systems. One possible way of doing this is replace vendor specific UNIX systems with lower cost Intel or AMD based hardware running either Windows or Linux.
I think that over the next 12 months we will see more and more companies moving mission critical applications off UNIX servers onto servers running Windows Server 2003. For a large number of these companies, these mission critical applications are based on off-the-shelf software that is available for Windows, Linux, and UNIX. As companies start to do these migrations, they need to ask themselves which operating system to run on the new hardware. As you look a little bit closer, it would seem to me that in an environment that already has a Windows presence (which most customers do), does it really make sense to introduce yet another operating system into the mix. If a company is trying to reduce the number of different platforms they support to reduce cost, introducing a new operating system that they would need to hire or train people to manage and maintain seems counter-productive. In contrast, most, if not all customers already have Windows Server administrators in house.
I think that 2005 is more likely to be the year we see companies migrate some applications away from vendor specific UNIX and onto Intel and AMD based hardware.
What do you think?
You say most companies have windows administrators. The whole momentum is shifting to linux. A system you are putting in now you will be maintaining in three years time. in three years time it will be easy to get linux guys but not easy to get windows.
also which version of windows.....only dummies will be migrating to windows.
Absolutely. In fact, according to Slashdot, 2004 was the year of Linux. And 2003. And 2002. And etc.
But seriously, THIS time it's for real! Microsoft is in huge trouble! You're going to lose your job in two months when Microsoft suddenly goes bankrupt because everyone spontaneously starts using Linux and FireFox!
Would be nice to save this page off in history somewhere, and next year when 2006 is the year of Linux, or two years away when 2007 is the year of Linux we can ask Runple Stiltskin and Jakeypoo what happened back in 2005, because according to them this is the year Microsoft suddenly loses 50 billion and files for bankruptcy; a mass layoff of all Windows admins happens; and companies that have years of development on a Windows platform decide to scrap everything and move to something totally new.
Exactly what Jakeypoo said.
My original comment was sarcasm (as of course was Jakeypoo's).
It becomes a moot point if people design things in platform-independent ways.
Runple, you say that as if running Windows or Linux is mutually exclusive, that an IT group must have homogenous servers... that isn’t too realistic.
Yes, there is a trend towards Linux, however few companies are going to throw out a perfectly good and working Windows machine just to throw in a ‘free’ Linux one. Much of the trend you speak of is adding Linux boxes to existing environments, not replacement.
In such a case, you still need Windows administrators around. I am very curious though... why do you think that in three years, Windows administrators will be so hard to find? While I do not consider myself easy... I know I will still be a Windows guy in 3 years, and will be for as long as Windows is a superior product to anything else available too me.
Also, do you call moving from one Windows server to another migration? I hope not, because far more people do that every day then do so to Linux... they must all be dummies then by your reasoning.
If anything Jakeypoo, Icontend that true year of Linux wont be until 2112... or more appropriately, 2113 (because of it being a full year) when by my math, the original version of Linux will enter the public domain (and only the original version then).
Forgive me Mr Blobby, but have you ever designed a medium or large scale system in platform-independent ways?
This last summer I spent a great deal of time working with an in house application that was written using the wxWidgets GUI Framework. In theory, using wx lets you write an application that can easily be built in Windows, Linux, MacOS. Unfortunately, due to plenty of odd little quirks of the implementations on each platform, it was a nightmare making the app work under Linux after it had been built under Windows (yes, the original coder would have spent plenty of time along the way making sure everything worked under both).
Designing for platform-independence is a wonderful theory, but rarely works, even when the systems targeted have rigidly defined specifications and standards that everything you write and use conform to.
I was going to mention that I was being sarcastic in my original post, but I thought that would be insulting because it was so obvious. I stand corrected.
Well you are correct, I work at a large Global Fortune 500 company, while I am not in the Linux Windows Debate. I am simply just stating what we have been doing Unix, gone, started moving out around 2000, this year all gone, Unix Work Stations been replaced with Windows. What were those Stations? $50,000 Sun Workstations for Catia and other high end CAD systems. Replaced with High Ram Dual Xeon IBM Machines Running Windows XP. Workstation cost, $6000 complete. Runs not only Catia but about 12 other cad apps with ease.
Still have an HPUX server and a Back Up server it may be around for sometime to come. But unix has gone the way of the wind, so has the Unix Admins. Really we do not want the hastle of the Unix Machines anymore or the cost of them cast of the Admins, when they are down they are down for hours with people standing around going hmm I don;t know did you try to recompile the kernal, maybe if you tried in a different shell. We also do not see Linux being any different, a lot of hastle to set up, deploy and the average user like a manger or something they are never going to learn Linux, they could care less they just want it to work.
Look at it this way, is Linux Easy for your Grandma to use? No, Can you Grandma Fix it easily. No. My 84 year old Grandma, never touched a computer before in her life. got a windows machine, She actually figured out on her own how to hook the printer up by herself. Now yeah the hard core people that live binary like Linux, the average person does not they just want it to work.
The other problem is Linux support. Look at linux help board. You go out and ask a question, "hey I am having a problem with my workstation it is doing this". On Microsoft board you get. 20 people going Oh Have you tried this, or here is a link to a Microsoft article explaining how to fix this. On a Linux board you get "Yo, j00 Lam3r, U Suck and Kn0w Nothing, g0 back To Winders newb"
Anyway, since 1998 every single year was promissed as the "linux year". And long before that, every year was also promissed as the "year that unix will took off as a mainstream operating system".
There are always this kind of horoscope news when the year is just beginning, and there always be. Even Steve Jobs did the same, by saing that 2005 is the "High Definition Video year", so no matter what kind of analysis we make, someone will say that this is the "[complete with your own thing] year" ....
I think Jeff brings up some interesting points. I work in an environment with mixed windows and several flavors of Unix. Linux isn't really a player here. Oddly, some of the middle managers think we need to bring it in house to investigate it. No real business reason. On the other hand, we have over 3K Windows servers in the environment and a mainframe.
Is Linux fine for many things? Sure. Does it make much sense? Not here. Does it matter? Sadly, no. The press time is what's getting looked at and causing folks to think that Linux is worthy.
My take? Best tool for the job everytime. If that happens to be *nix, then fine, use *nix. If it's a mainframe, then use a mainframe. If it's Windows, use Windows.
I think that in terms of implementation, every year is Linux year because the market share grows, so every year means an improvement over the last one. Also the community is taking Linux close to the end user side, with improved usability, and bla bla, bla. Also, as Windows was taking the most of the market they can only go "worse", so easily we could say that Linux goes better and Windows is loosing, but I don't think this is so easily comparable.