Link Details

"proprietary solutions" - absolutely not. You mention 10 years ago it was all proprietary, which is true, but it won't repeat. Simple reason is developers drank from the fountain of openness, and it taste great. No developer or small startup will ever go back to 100% company crapware produced and sold for loads of money. I remember paying $800 bucks for JBuilder back in the day, and let me tell ya, that's never going to happen again. As far as Java is concerned, developers will gravitate to what works for the littlest amount of effort (bang for the buck). Yes, companies will try to inject proprietary crap into anything they can get their hands on, but developers now have a much better nose to sniff out the crap and use only what works.

killerweb,

I appreciate your enthusiasm, but it will absolutely happen again and never stopped happening. Next time you get a chance to talk to a friend working for a bank, insurance company, large scale vendor, etc... ask them what they use. Most likely it's IBM software on IBM hardware using RAD at $4k a seat. It's not unicorn-inducing open source wonderfulness...

Vendor lock in never went away, it's just changing shape.

I worked for a very large GIA (General Insurance Agency) as the software development manager about 6 years ago. While we were using Weblogic at the time (I think they've since moved to Jboss) the company absolutely would not pay for anything like rational or jbuilder and we had a fairly small team (6-8 developers). This was before eclipse really got much traction. The only real options were Jbuilder (too expensive), Netbeans (too slow) and some small time players like Visual Cafe. Frankly, back then, most of us still used Gvim and ant and managed to write hundreds of thousands of LOC that way. ;-)

IOW, don't make sweeping generalizations--they just ain't true.

The Votes from developers like us, speaks louder than words. And yes vendors did change shape, from elephants to mice.

This is an absurd premise - there's more potential to build enterprise solutions on a fully open source Java-centric stack than ever before.

In our work place developers work on Windows. But our deployment platforms are AIX, Windows, and Linux. Linux is becoming the mainstream deployment target as we're leaving legacy AIX behind and only using Windows for .NET stuff. When I'm at home I work with our software on perhaps 64-bit Ubuntu Linux, or 32-bit WinXP, or Mac OS X. Because the stack I'm dealing with is mostly Java (and Adobe Flex for client), I'm able to move around on these various platforms at will (trying out different browsers, etc).

When I build apps to be deployed on servers - I hand off .war and/or .ear files to the folks that are responsible for that. I don't give them crapola php or ruby directories full of script files and hope they figure it all out as to how to deploy it. Instead they go drop my .war or .ear in a designated folder and drain-stop the app servers across the cluster.

Java continues to supremely kick ass for enterprise development. No platform ghetto lock-in ala .NET. Rigorous control of build configuration (Maven/Hudson) and deployment.

Until someone can point me to something that can top Java in regard to all my concerns (that Java is handily addressing for me), I'll quickly give them the figurative boot back out the door. So far I can't find a damn thing that can compete with Java in regard to the rich eco-system Java universe. It addresses so many concerns in so many facets of the software process from top to bottom, from beginning to end. And it's way more than the language itself (tools, libraries, frameworks, standards, open source, community, etc.). The likes of PHP or Ruby pale in comparison as those have very niche sweet spots. .NET could have the potential to rival the Java platform but, of course, is stymied to a single OS platform and innovation comes 90% from the single vendor that owns it.

Proprietary vendor lock-in - that is such a horse laugh premise as far as Java is concerned.