One could hardly imagine a more loathsome and pithy example for the arbitrary and capicrious means whereby pompous and pampered State bureaucrats throw into the void the money which WE in the private sector work so hard to earn — by serving the REAL wants and needs of other people, expressed by their voluntary purchase of our goods and services.

In this particular example, Vienna, they couldn’t finish their migration because of a software program that is not available for Linux. So the Linux platform is clearly lacking developers – despite the countless open-source projects that are available at Sourceforge. Microsoft invested into its developer community very early on, and the MSDN (Microsoft Developer Network) is a really very good resource for developers. Nothing comparable exists on Linux, and finding information can be quite difficult. If Linux wants to attract developers, then somebody needs to come up with a resource that is comparable to the MSDN. IBM comes to my mind – they already sponsor Eclipse and certainly have the resources and motivation. Yeah, there are thousands of developers that develop for Linux, but they are mostly developing for the kernel, libraries and system software. For the average developer, Windows still seems to be the platform of choice.

Note : I like Linux and have used it since ~2000 – dating back to somewhere around RedHat 5 or so. Windows has been my primary platform over the last few years though, with increasing focus on Linux/BSD/OSX as of recent as we start porting our product to Linux.

Anyway. Price is another issue. Yes, Linux is free – theoretically. But most companies that use Linux end up buying support contracts from RedHat, SuSe, etc., and if you look at the prices of those, you’ll quickly see that it ends up costing the same as Windows – if not more. I’m not taking Office into consideration which could change that balance, but don’t forget that Microsoft is giving significant discounts to universities and schools, so the price of Windows is often *significantly* lower than what a corporation would have to pay.

Stability is not so much an issue anymore either. I remember the Windows NT 4 days very well, when IIS 4 had more holes than swiss cheese. But they have evolved – at least on the server side – and I consider Windows Server a very stable platform. It’s also very cohesive, and includes a wealth of services (AD, DNS, Group Policy, Software Distribution, WSUS, Performance Monitoring) by default. Furthermore, all these services can be configured from the command line, with a GUI, or accessed through well-documented APIs. You simply can’t do all this with Linux unless you spend a significant amount of time setting it up. Even then, you have separate products, probably with separate interfaces.

The fact that there are so many Linux distributions available is an advantage considering the many choices, but for the user – especially a new user coming from the Windows world – it can be confusing. With Windows you have Server 2008 with a few different versions that are very similar. It’s truly amazing that the distributions figured out how to make all those different software packages work well together. Vista is a whole different story of course, and the many different editions Microsoft is offering are not well perceived in the community as we all know. XP was fine: Home + Professional – they should really go back to that. But anyways.

There is no question that the Linux kernel is a stable, rock-solid foundation. But the fact that so many components are involved, so many different vendors, makes it difficult to attract developers and sysadmins. The people that use Linux are a whole different crowd than those that use Windows.

I’m looking forward to the day when Linux will be more prevalent in organizations, and give Microsoft a run for its money. Microsoft is not destined to dominate the market forever, and the recent years have shown that it’s somewhat dropped the ball on the desktop front. The latest IE 8 is an inferior browser to Safari and Firefox, and Vista is more or less a failure that even I can attest to. But Apple should not be dismissed either – they conquered the entertainment front with iPods, and are using that as a base to get home users and college users converted to OS X. And it’s working. Before you know it, they’re going to provide services similar to Windows’s AD, and assuming they’ll do it right, customers will switch over. They’re definitely bridging the gap between Linux and Windows – positioned somewhere in the middle despite being a hardware monopoly.

To me, initiatives like the LSB are extremely important, and more attention needs to be paid to that. If Linux makes it easier for developers and sysadmins, then half the battle will be won. If not, then Vienna will just repeat itself over and over, because some ABC app will just not run on Linux. The Kernel, X.org and Gnome/KDE are just not enough.

That’s my opinion anyways, and I really should be working

Christoph, “Do not ask what TvöD can do for you, ask what you can do for TvöD!” This will be my favorite quote for the rest of this year.
I think it is not just that they wanted to see if they can move to Linux. Their main motivation was to put Microsoft in their place. In Munich, the whole thing started when Microsoft ended its support for Windows NT. They were outraged that Microsoft dared to push them to Windows XP. Thus I think this decision was mostly based on emotional rather than technical reasons.

And yes, it’s shitty paid

Despite that I rather see this as a “let’s see if we can do this and how it turns out”-projekt.

For an environment of that size, you’ll need very good professionals, no matter what OS you’re using. Yep, Microsoft professionals might be cheaper because the supply/demand ratio is higher than on the Linux side – but you’ll still need them.

And there are a lot of people in both environments without much of a clue. They’re dangerous, but the danger is the same on both side.

One other thing that i’d like to see mentioned:

Linux environments are more development heavy than Microsoft. If you want to setup central authentication on Windows, you use Active Directory. Every Windows professional knows AD and how it works.

On the Linux side, this is a bit different. You will probably go the LDAP route, but what kind of Schema will you use? How do you replicate your LDAP servers? etc.

For a big corporation this might be better because it allows the infrastructure to be developed exactly to business needs, but it also isn’t as “proven” as the standardized AD way.