Hey everyone! I'm writing this message in hope to get some serious advice.

My problem is:
After spending over a year programming for Windows I started asking myself over and over "Is that a right thing to do, developing programs for the platform, which future is uncertain?", "Do you think Windows is gonna continue to be a dominant OS on a market, for after MacOS X came out, attracting more people every day, and Linux becomes more and more popular among people and buisness organisations, and the quantity of Win32 programmers is growing everyday with a tramendous speed, will you have any proffesional value as a programmer on the future market?"
Someone I work with told me: "You need to program for Linux/Unix, because that OS prooved itself for the last 30 years and it's only getting more popular everyday. And that a Unix developer has the highest income rating, and the highest demand on the market?"
On the other hand, Mac based their last version of OS (MacOS X) on nothing but Unix, bringing lots of stability to their PowerPC users (wich also makes me thinking that Unix has one of the greatest potentials), but bringing a lot of dissappointment to Apple's developers, because they already got so much used to the programming for older versions of MacOS, that are completely different in concept from the new one.
All this makes me feel like to abandon MS Windows, stay away from Apple's Macs, and dig my way through the mud of Unix.
But who knows how is everything going to turn out. May be Microsoft is going to come out with something new tomorrow and get everyone back to it's dirty buisness of making profit for themselfs. May be Apple is going invent something different and fiil up the market shelfs, or some other weird and never known before company will create some kind of platform independent Java based OS, with a power of its interface, that will allow you even to write a driver for the platform..
May be among all the dirty wars between different OS manufacturies the only answer for a programmer who wants the ability to write good applications, that will serve a usefull purpose for people for a long time in future, is to deal with Java technologies without having a headache of thinking about everyday's changes in OSs requirements? That would be great! The only problem is that on the current stage Java is a very weak and slow language, that doesn't let you do as much as you can do with knowledge of C++, assembly, OS API and some other advanced technologies. Yes, please don't get me wrong that I don't like to programm for Windows, I actually love Visual C++ 6.0 and think that it's one of the greatest programs I've ever seen in my life.
So after thinking of all these problems I decided to make some research, analise some statistics and may be come out with a better idea of what to do. But right now I'm, desparate in searching for the answer, asking you to tell me what you think of all that? I hope to get very serious and conciderable advice, PLEASE don't send some quick replies like: "Hey Do the LINUX". I will aprecciate some well analised discussions.
Thank you very much for reading this and if interesed I'll be glad to see the response from you, and who knows, I may be see you in the next forum? :)

Depends what sort of industry you want to work in. Do you want to make GUIs, work on databases, manufacturing, financial etc. Pick the industry you'd like to work in and see what the dominant OS is from the job adverts.

It may be Linux in film studios, Windows in accounts, sybase in databases, Java on web pages, MAC OSX for publishing etc. What happens next is anybody's guess but whatever industry you choose will have its own specialized programs and its favourite operating systems.

It sounds like you're a win32 programmer and you're worried that your marketability will decrease due to influx of new win32 programmers. And you're hoping to gain an advantage by guessing where the market will go, and moving there before the gang.

Well, if you guess correctly, more power to you! But I don't see how anyone could make that kind of prediction with any degree of accuracy. Certainly there are better operating systems than Windows in a lot of ways. It's obvious that you can't predict what your marketability will be based on knowledge of the superior operating system.

There is room in the world for more than one operating system. And you can probably make money with either choice. However, there are going to be some more marketebility questions that you can't answer. Your sales will be limited if you choose the wrong OS for your product. And if you choose multiple OS's to release your product on, you'll have to deal with the headaches of that. Unless you're a niche product, you'll most likely sell more windows product than your Unix, Mac, Amiga, Linux, DOS, and NetWare combined. Check some companies that have had products on multiple OS platforms, and see which ones they are still supporting, and which ones they sell the most of. Ask them if there is a change in the trend.

Don't be convinced by arguments like your friend's, "You need to learn Linux because it has proven itself over the last 30 years, and is only becoming more popular." You could say the same about Unix, Mac, and Windows. Certainly Linux is a great platform, but there are some inherent problems with it. Same with all the other OS platforms.

The other issue is that once you learn systems programming on one platform, it's a lot easier to learn another one. Some of the specifics change, but most of the major concepts are the same. Certainly it would take some time to become proficient in the new environment, but I think you're overestimating specific platform expertise and underestimating general programming expertise. I say pick one now and get busy. If you need to learn another one later, you'll know two. :)

What's the point of arguing whether one platform is better than the other? Which one are you planning to be using? If you're an all Windows operation, staying win32 might be good enough for you. If you plan to port some of your servers to linux, bsd or whatever, you might look into unix programming.

Ask yourself, how does this affect me and do I think I am going to need it? Then ask yourself if learning a new system is really that bad an idea. The more you know the more flexible your skills. You could change jobs and find out that the new company doesn't use Windows or unix at all. They use AS/400.

My point is that I don't see any reason not to learn something. Download Red Hat or Slackware or Mandrake or whatever, install it on an old computer or dual boot and see if you can reinvent some of your windows software using qt or gtk. It doesn't hurt anything to try and it may give you a nice portfolio to show to a prospective employer (or your current one for that matter).