TwoSheds

+ Finding or creating suitable functional interface for passing the function

How about this document http://www.ecis.eu/documents/Finalversion_Consumerchoicepape...

It's fixed now, the rule is that one of the team must be a citizen of the country the team represents.

That's great news! Maybe I'll find some use for my N950 soon.

It's not really a singleton if it returns an instance per session instead of a global one.

NetBeans is open source. At least OpenOffice and Hudson have already been forked. Are there any obstacles (name has to be changed of course) for forking this project as well?

Apparently ISO is not one of them, OOXML required zero implementations

XP has the "windows rot" problem. After it has been in use for a couple of months, it slows down significantly and especially booting it takes forever. Some people claim that they can avoid this but I have never seen any hard evidence :)

Good point. And I don't think classpath exception would apply here. I'd be interested to hear if someone has actually seen this source code "in the wild".

AFAIK the code which makes the Swing UI use native OSX (Cocoa) instead of X is Apple proprietary. I wonder whether Apple will now give that to Oracle or release it as Open Source. Otherwise we're going to be stuck with X Server and ugly UIs for a long time :-(

Nope.

If Oracle decides OSX is important enough platform, this might be even a good thing.

In my experience (last 5 years in MegaCorporations) the problem is that the people who are actually hands-on managing the offshoring projects are different from those who make the decisions. The guys who decide to offshore are usually at least 2-3 layers of management above.

The urge to offshore software development at that level is strong, and the management sees it as a megatrend which "we'll have to follow because everyone else is doing it".

The feedback loop is broken, the decision maker who offshores a certain project is no longer there when a couple of years later it's clear that the whole thing was a disaster. He's busy offshoring some other project.

Scala needs the JVM (well, at least it's main implementation does).

There were at least these cases which seem trollish:

MS vs. TomTom

http://www.zdnet.com/blog/microsoft/chewing-the-fat-in-the-m...

MS vs. Salesforce

http://37signals.com/svn/posts/2341-microsoft-patent-trolls-...

There was a survey about this (among other things): State of Clojure:

http://muckandbrass.com/web/display/~cemerick/2010/06/07/Res...

Quite a few Ruby and Python people trying Clojure out (much more than Lispers).

You've got a point, but I think final class modifier which prevents extension often hinders Unit Testing and never adds much value.

I have 600 MB stuff there, and it certainly beats my corporate Outlook in speed...

Last time I tried BitBucket it just didn't work over my employer's http proxy. Googled it and it was some unresolved issue back then, lots of people had the same problem.

Exactly. I work in medical hw/sw and I can tell you the amount of bureucracy is insane and everything is orders of magnitude harder than any non-regulated business. Open Source itself is not really a problem, we use a lot of it (os, libs, etc.).

Unfortunately?

Isn't this a bit similar than what GWT does? GWT requires Java which is clunkier than a modern dynamic language, but still...

It runs the whole Android system including Linux kernel. It doesn't perform that great on a desktop, it wouldn't probably fit into iPhone at all this way memory-wise. If it did, it would be extremely slow.

And you never have to debug and fix your bugs? That takes zero time?

Yet another article which claims that development is faster without Unit Tests. I used to be a cowboy-coder in the past for years, and now I always do Unit Tests. My experience is that it's always faster unless I'm just doing small API exploring which I throw away. Without tests I always end up debugging longer than I would have spent if I'd written the tests.

And the future maintainer has much better chance of determining whether whether certain behavior in my legacy code is a bug or feature - desired or accidental.

I don't think these guys have done any serious TDD or Unit Testing projects. They have just tried to write some tests and then they have given up because they just thought it's too hard and their tests were probably huge, testing too much and not really maintainable.

Yes, I agree totally that S60 as a programming platform is horrible. I hope Nokia has learned the lesson with Maemo. Because it's based on a Linux distribution which is not that far from your standard desktop Linux, maybe they just aren't able screw that up too badly. To Nokia's credit they are not the open software development freedom-haters that Apple and maybe Palm now too have become.

Why does everyone think Droid is the only hope? Nokia's Maemo-based N900 looks promising as well.

Their latest Symbian/series60 phones (e.g. E71, E75) are also pretty decent nowadays, although I would compare those to Blackberry rather than the iPhone.