In fact the second case is somehow wide spread in our codebase as we need it as a part of type enforcement.

Ok, please note that I had carefully said that JVM have no support of Generics. The Generics encapsulation are checked at compiling level. This approach keeps binary compatibility (so is for good reasons) but in fact you can take a “generic” class and replace with one non generic. Also, without wanting to enter in details, I say that this is why .NET 4.0 adds support of covariance just because the .NET 2.0 offers support for Generics. Java have no such problem as it does not “care” at the bytecode level.

So in Java, ArrayList is translated at bytecode level as ArrayList, when in C#, List is in no way compatible with List, so using other type instead will make at runtime level to nothing happen.

The type enforcement can be done in a lot of ways, there are adnotations (the preferred way to do it on Java 5 and up, which adnotations are much better implemented in Java), but is not based on Generics. In Java also a complex generic code goes to a lot of conversions with generics, so it gets down in speed, when the equivalent .NET implementation of generics, will improve performance and memory consumption (excluding the fact that you functions associated with every type).

At the end I want to say a thing: both C++ and Java have their own strengths. C++ is snappy, no JIT occur at any given time, it integrates seamlessly with any C or C++ library, and with Qt things goes pretty decent. I am sure that is a very good reason to implement Photoshop Elements in it, or the best use case IMHO is for phones because no JIT-ing is involved. Java also it have a lot of professional grade tools. And at least for web development using EE frameworks is great. A bit big, but is great.

Anyway, C# at least in combination with word: Desktop (and you may note: only on Windows) is pretty great competitor. Language/VM wise is a powerful implementation, let’s say comparable with Java. Also, SharpDevelop (I don’t care about full priced VS) offers as the last stable release: code quality checker, code coverage, unit testing integration, integrated profiler with minimal overhead, one build tool, versioning control integration, visual designer tool. This as features at least is the same with NetBeans on desktop, but yet I think that Windows Forms looks better than Swing on Windows. And at the end is: on our project, was better fitted than other platforms. It does not mean that other were bad, but in our evaluation, .NET as development platform was the best on that given moment.

http://en.wikipedia.org/wiki/Generics_in_Java

In any case, I hope you are keeping your options open.

The single compelling option as platform was Java, but JVM have no generics support and as CAM/CAE parameters and parametric modelling programming may break very easy your data, so the second technology was not preferred.

So here was only about picking a technology that few developers (to be more precise: two) can create under an year on top of a CAD opensource framework (OpenCascade) a full working concept of a CAD/CAM/CAE application. Linux WAS considered as an option but there were many small limitations (and the biggest was the build time and harder to setup environment for some, as QMake was fairly easy to use) that makes us run away.

Can you imagine other framework than Qt or Python to make a desktop application that look almost integrated in your desktop? Gtk and GLib OO programming is a joke. Also, if you put the IDE power on the platform as a PAID developer that need to feed family and to invest in hardware to contribute back to opensource (as most software works today), even paying an IDE in a closed source environment I think is not have critics as long as I do have productivity.

As matter of fact I use SharpDevelop and I test small details on Visual C# Express.

It’s easy to speak about freedom as long as you consume it, but I know that I have to write a highly complex parametric code, that generics are mandatory and I remain with two options (Java is not one of them) on Linux: Mono and C++. If you put that C++ for large projects translates in 1 hour build time, I think that if someone will come and say: take your project and rewrite to use Mono, I would say: I will give to you my help, when you want to donate time to do it?

The last thing that I complain on Linux is that I bought an ATI board (because AMD was pretty opensource for some time) and even now it’s OpenGL does not work smooth. Is a 4850 and is an entry level professional card. I use for a “single game”, Vista Aero which works almost without a glitch.

What I want to point here is the fine border between freedom and hypocrisy and persons dictates over their personal preferences to write a thing in this or that, but it does not match when we talk on real life developers that have some time to spare, and if tools are wrong, they should better use a SharpDevelop (let’s say) and at the end to port it to Mono, than to use makefiles and other abnormalities like C/GLib only because an user would like for their fun to consider that this framework is their first love, but not doing programming with it, but only making fun of developers to use that framework. Look on Gnome 3 that will work for at least one year to remove deprecated symbols and technologies. Sounds as fun, isn’t so?

Well, even if it really happens it is taking so much time that Microsoft will do a lot of cash in their way down.

Am a linux user since the late 90′s and I use it as my default lilo/grub boot option since 2002 or so, but I still can’t recommend it to people too used to Windows and/or too dependant on apps that ONLY run in Windows (and that have no opensource “alter-ego”)

Microsoft monopoly has too much inertia to be removed.

In the news:

“The situation puts Microsoft in a quandary. If the company lowers the price of Windows 7, it could hurt revenues and profits. If it keeps the price high, PC makers might bolt to alternatives, such as the free Linux operating system.”

People keep blasting what year is the year of the GNU/Linux desktop. I have to say it was over 3, close to 4 years for me. Every year is the year of the GNU/Linux desktop. Growth is far faster and viral than people realize and that it’s momentum is an inexorable, unstoppable force.

My thought is the next OS to truly be revolutionary will then win…weather if that is Windows, Linux, Mac OS or some new OS being developed. It would have to really be revolutionary, something that people will want to go to. It won’t matter if its free or cost money, if the next revolutionary OS (not just evolutionary) has something people want and will make things easier. Maybe it’ll have excellent and seamless voice recognition, maybe it’ll have an easy to use touch screen or virtual screen and maybe it’ll have both voice recognition and a virtual screen? Maybe it’ll be controlled with the brain (may sound silly but people are working on it). I think as long as we’re using a mouse and keyboard, Windows will be the top OS and thats because its backwards compatible. As long as Microsoft keeps some sort of backwards compatibility, its not going anywhere. Unless a linux distribution can give a high amount of easy compatibility with Windows programs, Windows will be on tip. This new OS will have to make people say “screw backwards compatibility” and want to start from scratch. Mac OS isn’t doing that and Linux isn’t doing that.

Another thing that can maybe change Windows from being #1 is right now the computer is something we use in our house and work, we use it for fun and for business, but for the most part its an appliance. For another OS to take over they’ll have to make it a seamless extension of our lives. Meaning weather if we’re in our house, car, work or anywhere in the world we can access all our information, programs and games. Its not about transferring media from your home PC to your PDA phone, but to just have it work without transferring. Being able to go on a trip and to be able to get stuff easily from your home or work computer seamlessly, quick and easily. Having this work without even having to think about it, it just does. No need to have several computers in your house for each person who lives there, just have one that everyone can share anywhere in the house on several different medias weather it be a terminal, or handheld device.

If either of these two ways…or both come out then Windows will have a problem. Maybe it’ll be a single computer in your house and you’ll be able to access it from anywhere, maybe it’ll be what Google talks about, apps that are run from the browser and it doesn’t matter what OS you use. I’ll be light and simple and you can have it on your PDA phone, and home server and they’ll be able to communicate with each other and everything that is on one will be accessible on the other. Work servers will run off of it and with permissions you can easily get of your work off those too.

Thats my theory and thoughts on it and no 2001 memo is going to make me think differently. Sure MS is worried about Linux and Mac OS, and anyone else thats thinking of an OS. Because any year some company can come out with an OS that will make people want to switch, but its going to have to be a great OS and we’re going to have to be WOW’d. Nothing has come out yet that does this, when will it happen? Who knows. Will it be Microsoft or some other company that develops it, that is yet to be seen.

The winning OS will be the one that have all bases covered. A OS that a workplace wants to use as it knows it’s workers use it at home. A OS easy to use and fast. An OS that doesn’t restrict the user.

When Bill talks about Window being an appliance in the product sense, he’s really nailing down Window’s restrictions. We must play on that.

A great slogan for Linux/gnu would be, “Everything is happening on Linux”. Community and Hobbyist themes are our best shots.

You should want freedom for its own sake and this is why M$ is evil.

We fundamentally disagree. I don’t value freedom in itself. I just value what can be done with it. With software, freedom enables an atmosphere where wonderful things can be done. With land ownership, freedom enables the draining of wetlands and all sorts of evil. Some lack of freedom is necessary if we want to live on a wonderful green planet.

It seems quite relevant to note that there are no negative externalities from software development, but there certainly are from real estate”development”.

We need other creative ideas to allow Windows to match the viral nature of Linux.

This is ultimately the reason why Linux will triumph. It’s not costs for the consumer. It’s not even freedom. It’s creativity. Microsoft is a large bureaucracy that squelches creative ideas. Linux lets the thousand flowers bloom and thus Linux evolves.