Java and .net

[Luca Leonardo Scorcia]

While I agree that you cannot compare Mono and the .Net Framework, you're wrong in some assumptions: .NET 2.0 support is planned and under heavy work and you can as well run applications built using Windows.Forms, just they don't feel as smooth as native (GTK) apps. And it's going to be better with newer releases. Luca The Price of Freedom is Eternal Vigilance.

[Chris Losinger]

Marc Clifton wrote:

Keep in mind, teaching Java replaced teaching BASIC

i had 3 years of Modula-2 when i was in school, and no BASIC at all :( Cleek | Image Toolkits | Thumbnail maker

[jith iii]

Ryan Roberts wrote:

jitAtran wrote: some C or C++ Which are far more difficult languages than either .NET or Java. C has pointers and C++ has pointers and mindblowingly complicated and powerful sideffects of its template mechanism

Sorry dear ...I must have changed it into some basics of C or C++.Generally in india even if you are studying for electrical engineering or mechanical engineering you will have to learn C or C++ . -- modified at 10:02 Monday 1st May, 2006

[Dan Neely]

Luca Leonardo Scorcia wrote:

you can as well run applications built using Windows.Forms, just they don't feel as smooth as native (GTK) apps.

Can you run them cross platform? It was my understanding tha gtk was needed if you wanted to target a nonwindows desktop.

[Luca Leonardo Scorcia]

Yes, you can. Of course, many advanced features are under development, but you can easily take a dialog based Windows.Forms binary and run it under mono. They have rewritten the entire Windows.Forms using GDI+ calls. Slow, but works (mostly). Check out the screenshots on the mono-project web site. Luca The Price of Freedom is Eternal Vigilance.

[jith iii]

I have been hering about Mono and DotGNU for quite a long time.But nothing really comes out into the developer's arena. Also mono is from novell which owns Suse Linux.So what would be microsoft's real strategy. Evidently projects which aims platform independance are choosing java.Is "windows" killing .net. -- modified at 10:41 Monday 1st May, 2006

[Paul Conrad]

David Stone wrote:

I hate it. Java sucks. And Java's generics suck harder. And in general, Java just sucks. Give me C# and .NET any day of the week.

You've got that right David. I tried out Net Bean 5.0 from Sun over the weekend and back to VS2005 I shall go :) Paul

[Giles]

C++. Java, is for girls. .NET is for people who want to get the job done quickly.

[David Stone]

heh. Glad my late night rantings were of some use. :rolleyes:

Oh geez... the forum keeps spinning... you'll take care o f it i'm sure, c'ause ... yeah, i neede this. *cough* anyway good job finding the bug.
-Shog9 on...a Firefox bug.

[David Stone]

computerguru92382 wrote:

I tried out Net Bean 5.0 from Sun

Oh wow. I'm not even that adventurous. I'm using Eclipse 3.2 RC2 to code...and while Eclipse is a decent IDE (It's got a crapload more in the way of refactoring. Such fine grained code style prefs too.), I still have to code in...Java. X|

Oh geez... the forum keeps spinning... you'll take care o f it i'm sure, c'ause ... yeah, i neede this. *cough* anyway good job finding the bug.
-Shog9 on...a Firefox bug.

[Paul Conrad]

David Stone wrote:

I'm using Eclipse 3.2 RC2 to code...and while Eclipse is a decent IDE

I've had to investigate Eclipse for my Master's Thesis and it seems pretty alright. PJC

[Mike Dimmick]

Personally I still think pointers are actually easier to understand than C#, VB and Java's reference semantics. There's a syntactic difference between manipulating an area of memory ('object' if you prefer) directly and manipulating one at arm's length, through a reference. Most pointer-based languages do have the problem that you could be manipulating variables of a given type either directly or through a reference, while C# and VB make a clear distinction between value and reference types (struct and class respectively in C#). The CLR allows you to 'box' a value type to use it as a reference type, however. Having said that, I learned pointers using Ada, which does not have any syntactic difference at the point of use to indicate use of a pointer. C's problems typically come from the ability to create highly convoluted declarations - these are not necessary, it's perfectly possible to write clear C code - and explicit use of memory deallocation routines. However, a correct C#, VB.NET or Java program must make use of the Dispose pattern, for all but trivial programs. In C++ one can mask most use of the memory allocator and deallocator, and indeed any other resource deallocation, using the Resource Allocation Is Initialisation pattern. This pattern is so engrained in C++ developers, generally, that they're not even aware it's a pattern. Simply, you use a class to manage the memory or other resource; the class destructor frees the resource. If an object of such a class is created on the stack, the compiler automatically generates code to call the destructor on exit of the block. If created on the heap with new, the destructor is called when the delete operator is used. Diligent use of this pattern normally leaves very few objects whose lifetimes are managed with explicit new and delete statements. C and C++ both have pointer arithmetic, and give no warning of writing outside the bounds of an array or other memory allocation. The former is not required in writing a program - pointer arithmetic can always be substituted with array indexing, and compilers are generally very good at converting back to pointer arithmetic if possible. The latter requires some discipline in tracking the amount of space allocated; again, in C++ you can build a class that manages this appropriately, and the Standard Library and other commonly-used libraries have classes that will meet most users' needs. Stability. What a