Post a Comment
Yeah, all roads lead to redmond, except those leading to Java, Linux, the BSDs, OS/2, Solaris, OpenVMS, etc.
(Python runs on these platforms too.) Personally, I think IronPython is a great thing for people who want to go the .NET route, but still leave themselves a decent exit strategy. Sure, your code may end up relying on the .NET libraries, but at least your core logic will be written in an extremely portable language.
You really shouldn't underestimate the degree to which library dependencies make it unprofitable to retarget a codebase that wasn't designed to accomodate transitions to other platforms. Coupling your code to the class libraries available on the CLR isn't meaningfully different from coupling your code to the class libraries available on the JVM, or writing your program to depend on frameworks available only on Windows.
Dynamic languages for the CLR will always suffer bidirectional interop hurdles with the static languages that dominate the platform. This will limit the usefulness of Python for developing class libraries on the platform, relegating it basically to a position of gluing relevant frameworks together to make applications. This really only encourages coupling Python code to library code that cannot be relied upon to be portable in the sense that it is available on platforms other than .NET.
It should also be pointed out that IronPython performs better than CPython because it can compile directly to .NET bytecode that does not require the Python interpretor to run. That's another reason one might want consider IronPython.
But also, when you think about it, really this just amounts to porting Python to another platform. Just like Jython ports Python to Java (and has similar benefits to IronPython in that it can compile directly to Java bytecode that does not require the Python interpretor to run).
>>It should also be pointed out that IronPython >>performs better than CPython...
A far as I could see, it isn't faster than Cpython. What's more, in most cases it is much more slower.
However, I've never done any intensive testing. I just tested the speed of very simple code snippets, and they always come up being much slower in Ironpython than in Cpython.
This is not to criticize, IP is still in alpha, and performance optimization is always the last thing to do when creating a language implementation...
Mono has pretty much full support for .NET, except for Windows Forms which is pretty lackluster at the moment. On top of that it's got loads of extra (especially Gnome) class libraries. Some languages are even developed especially for Mono (like Nemerle for example), .NET compatability is just a biproduct in those. Mono is most certainly not a "hack" or whatever you meant to imply with that word.
http://lists.ironpython.com/pipermail/users-ironpython.com/2005-Oct...
(True == False) works now
Lol, yeah, if you want a language that is like python but isnt limiting itself to the bounds of python, try
http://boo.codehaus.org/
As nice as it is that there are so many languages that can be used with .NET, I have to agree with those that say you should stick with "native" .NET languages. I would even avoid second tier languages distriubuted with .NET, such as VC.NET and J#.net. Realistically, I would stick with either VB.Net or C#. VC.Net is a kludge (since C-type languages will always struggle between ANSI C/C++ and the CLR with its Garbage Collection, etc). J# is the last way I would want to do Java development. It's nice to know I can do Ruby.NET, but I am not going to be able to create an ASP.Net application with the code-behind in Ruby. And for me, that's the end of the story.
> why could you not use ruby in your ASP.net page?
You should be able to, theoretically. Visual Studio .NET 2003/5 won't help you do it though. If you really wanted to do it, you could write your code-behind in Ruby, compile it to IL (.NET assembly .dll) then write your .aspx front-end. With .NET, once the code is compiled it doesn't matter what language it's written in.
I don't know, but it may be impossible to write inline code in your .aspx page using Ruby until a Ruby .aspx compiler is written. That's a different thing than just a ruby compiler, which is already well underway. So I think you could write your code-behind in Ruby, and any inline-code in your .aspx page using C# or VB.NET or another language with an .aspx compiler.
IronPython is very interesting for GNOME/win32 portability (for those who don't like mono for whatever reason). As regards language features python is pretty well equipped and using the gtk (+ ported gnome) stack it is imaginable to create applications that don't use any .NET classes.
I believe this post is in error. It describes the IronPython project as being a programming language for the .Net platform. I believe the description you are looking for is, "Microsoft Visual Studio's .NET" or perhaps even Microsoft's .NET platform. My understanding is that IronPython has stopped working with any other platform since Microsoft purchased the head coder.
Last time I tried IronPython (I think it was the first release based on the .Net 2 beta) it wanted to contact an MS site each time I started up.
I don't know if that's the .Net library or IronPython, but it leaves me feeling rather uncertain. With all the stuff being written about 'Trusted Computing' do I want my fumblings with .Net to be broadcast to the world? I don't think I trust the .Net/IronPython combo. It may all be entirely innocent, but if they want some usage statistics they could just ask.
robin at reportlab



