Barney's Blog

Blog archive

Doug's Mailbag: What About Java?, Less Is More with Apps

Readers share their thoughts on Java and Microsoft's recent decision to stop supporting Java Virtual Machine:

Perhaps it's not as simple as whether Java is a winner or not. The current long-term economic contraction coupled with the recession results in much lower IT budgets. Lower IT budgets encourage CIOs to be less willing to 'burger king' it, and to instead do with off-the-shelf products or postponed efforts. Net effect: less development overall, Java or not.

Java is a cool technology, but so is .NET and, above all, good ole C++. Java created this illusion that it was going to integrate the whole world -- coffee makers, microwaves and all our daily technology would become some Java application. Which insinuates that it would be connected to a remote server somewhere, making the world one big smart client/cloud computing paradigm.

However, the problem isn't the idea, but the hardware and the means by which we compute. The ideas behind our advanced 'beyond the Web browser' Java are really the ideas for nano-technology. And we wouldn't have the same computing means as what we have with Java. Perhaps we'd do some development using the same methodology as we do now, but the nano-technology would be the secret, not the language itself. All a language like Java is for is to provide a more fun and accessible abstract layer to C/C++. If those languages can't do what the paradigm suggests, then Java isn't going to be much help either.

I cringe every time a business unit wants to deploy a solution that requires a Java client. Which JVM version does this vendor require/support? It seems that with each upgrade or new app deployed, we spend an inordinate amount of time testing different versions of JVM to find the one that causes the least number of problems across our (relatively small?) portfolio of Java apps.

Which do you prefer: fully featured apps, or slimmed-down ones? Doug asked, and readers answered in favor of the latter:

I hate apps with zillions of features because you can never find what you really need. Less is more. I developed software, and there was always pressure from some customer for one more feature that no one else will ever use.

Many of our users only need functionality that slightly exceeds the free Office Viewers (just add the ability to do basic editing). We would love to see slimmed-down, simplified and much less-expensive apps. The biggest thing preventing us from considering another solution? Almost all of our third-party solutions only provide integration support for MS Office.

And finally, two more readers debate which will reign supreme in the future -- fat clients or thin clients:

In our shop, thin doesn't even include a browser on the client. We've been majority thin client for years using Wyse Blazer/ThinOS devices connecting to 100 percent Citrix XenApp desktops, and the percentage of thin to fat increases every year. If Wyse and Citrix get multimedia working well (especially Flash), the percentage of thin will jump even higher. In a few instances, we use "chubby" XPE clients (mostly where someone has to have local Flash and multimedia capabilities) and we've recently started deploying XPE-based laptops for those with remote access/traveling needs. We're waiting for someone to provide XPE on a netbook with a simple docking station for keyboard, monitor and mouse when in the office -- that would be a really sweet solution for many of our execs, salespeople, etc.

From my perspective, the future is primarily thin for those who rarely leave the office, and probably "chubby" for most other mobile users and home users. Fat is reserved for servers and extreme power users (and those who require disconnected capabilities).

Here are my two cents for the future (might be worth a bit, at least at the restaurant at the end of the universe). In 10 years, we'll have wildly faster CPUs, bigger hard drives, bigger flash drives and Internet connectivity that's fiber. But once we get fiber at the home, I doubt it will get faster by much. The Internet is big cash, and the providers will need to take a break and make some cash on the existing base. On-the-fly compression may help, but who knows? So, faster CPUs and new flash drives -- maybe even ones that replace hard drives.

That tells me: Big, fat clients that only use what connection they need. It still may be cloud, but do you really think Fortune 500 customers are going to want their information in the cloud? Maybe in 50 years. There will be cloud apps -- those that focus on minimal needs, like Twitter -- but AutoCAD? Not for a long time.

Tell us what you think! Leave a comment below or send an e-mail to [email protected]

Posted by Doug Barney on 07/13/2009 at 1:16 PM


comments powered by Disqus