Slashdot is powered by your submissions, so send in your scoop

 



Forgot your password?
typodupeerror
×
Programming IT Technology

Shared Source .NET Ported to Linux 52

bjepson writes: "Shaun Bangay of the Rhodes University Computer Science Department has released a port of Rotor for Linux. You can find more details, including a download, at the O'Reilly Network."
This discussion has been archived. No new comments can be posted.

Shared Source .NET Ported to Linux

Comments Filter:
  • by zulux ( 112259 ) on Thursday June 20, 2002 @04:13PM (#3739025) Homepage Journal

    From the licence attached with Rotor:

    "You may not use or distribute this Software or any derivative works in any form for commercial purposes."

    Rotor on Linux is as about as usefull, to me, as a Corvette is on Mt. Everest.

    Except that the Corvette is cool, and Microsoft .NET is a spasdic mess of Windows APIs, p-code, neutered languages and FUD against Sun all wrapped in a license that sucks.

    • A spasdic mess of Windows APIs?

      Maybe you should try using it before you try to say something about it. Using the .NET framework you never see the Windows API - you see all new stuff. Pretty nice stuff too.

      Just as example implementations of a whole ton of core CS concepts and an example implementation of a working VM the Rotor source is useful. Go read some of it.

      - Steve
      • I have to agree, .NET is simply the best tool out there for development (course it's also the newest to be fair). And from what I've seen, Java(a.k.a. Write once, debug everywhere) is more of a spastic mess than Visual Studio 6 ever was.
        • I have to agree, .NET is simply the best tool out there for development

          Possibly - if you don't mind that any large .NET app is locked into a single operating system family running on a single hardware platform with a single language vendor.

          You say .NET is the best. Sure that may be true for limited values of best.

          In fact .NET is soooooo goooood that there are:

          No operating systems written with .NET
          No comercial games written with .NET
          No office suites written with .NET
          No drivers written with .NET
          No large apps at all with .NET except .NET itself

          (perhaps I'm wrong and someones cobbeled somthing together in these catagories, I don't keep active track)

          Come back later when you've actually written somthing non trvial with .NET and we can discuss it.

          • hmmmmmm Yeah.... Nevermind the fact it's been out for a couple months now... Oh and by the way, the IDE for VS.net was written in 100% C#, the compiler is not. If VS.net isn't a "large" app, I don't know what is (6 CD's) Maybe you should come back when you can show me a better tool for development instead of boo hoo-ing...
            • Maybe you should come back when you can show me a better tool for development instead of boo hoo-ing...


              Come back when you get a few years experience. You will then realise that...

              1) Any language that is only, as you state, "a couple months" old has bugs...
              2) Now one developemnt enviroument can be the best at everything.
              3) Good programmers know mutiple tools and can pick and choose amung them.
              4) Your favorite tool for your favorite type of app can quite often be the worst tool for a diferent type of app.
              5) Non-crossplatform tools hurt you in the long run.

              I'm not imppressed at all the the IDE for the new Visual Studio is done in C#; C#, and more specifically, it's libraries were designed from the ground up to make 32-bit Windows GUI apps.

              Get back to me when someone writes a device driver in C#, then I'll be impressed.

              There's nothing wrong with C# if you're making quick and dirty 32-bit Windows apps, but to say is the 'best' at everything, especially when it won't even work for embedded applications, betrayes a lack of experience.

            • My answer to all of this nonsense is the following: Every programmer has his or her own preferences. In the end, does it really matter? Code written in assembly or binary might impress us, but otherwise, who cares? Freedom to choose is grand. Let's face the facts, and see that the only reason there is a dispute over .NET is that it is written by a company that is apparently trying to create a monopoly in nearly every market related to technology that it can. Just leave it at that.
            • I've recently upgraded to VC++.NET I did so hoping that some of the issues I had to v6 would go away. I was pleased to learn that they did.

              I was horrified to learn that they had been replaced by far more severe bugs. I will list a few of the ones which perterb me more.

              - Cannot dock the properties window, the UI will hang when I switch from viewing a resource to a source file.

              - Intellisense is better than in v6, but doesn't resolve properly if you create a typedef to a parameter type

              - If the complete definition is not available for a class when you wish to pass it as a template parameter, the generated code will have no knowledge of that type. This is a problem if you define a CAutoPtr for some type for which you provide a forward declaration, but haven't included the header for that type beforehand. This is a bit of sneaky bitch, because the compiler will only create a given template once, and if it generates the one without the declaration, you're fsckd.

              - Finally, it just up and locks up on me. Usually after a build, I think. I've really started getting in the habit of saving. (I hit ctrl-s ever time I pause to think, lest the system be unresponsive the next time I touch the keyboard)

              So obviously C# is not the silver bullet. Neither is Java, mind you. The only cross-platform language is Perl (or Python, for those who's minds work that way)

              I can show you a better tool for development, it's not hard...there are so many. Apple gives on away. MetroWerks makes a schweet product. Even Borl...Inpri...Borland's stuff is better.

              Maybe my criteria for an IDE are different than yours.
              • After spending a grand on VS6 Enterprise, and a grand on Delphi/Kylix I refuse to spend a grand on .NET when VS6 is still useable....but I do 95% of my devel on Linux now anyway. If there exists a cost-effective .NET IDE for Linux, I'll get it. To even LEARN .NET I have to spend at least $500...and unemployment sort of makes that an impossibility
            • Oh and by the way, the IDE for VS.net was written in 100% C#, the compiler is not.

              Visual Studio .Net was not written in C#. It was written in C++. A lot of the .Net Framework is written in C# though.


          • In fact .NET is soooooo goooood that there are:

            No operating systems written with .NET
            No comercial games written with .NET
            No office suites written with .NET
            No drivers written with .NET
            No large apps at all with .NET except .NET itself


            s/.NET/perl/g
        • by pmz ( 462998 ) on Friday June 21, 2002 @09:25AM (#3743394) Homepage
          ... .NET is simply the best tool out there for development....

          This is true only if you do not care about risk. .NET is ruled by Microsoft; what happens when Microsoft is 1% of what it is today? Don't argue back with those "Microsoft is God and is immortal" type arguments, since Microsoft is a company like any other.

          If you work on an application that is really important to your company's long-term health, and you want your company to be successful regardless of Microsoft's success, then .NET is simply the worst development tool out there. A well-designed C or C++ application (i.e., modular UI) provides much better risk mitigation, since mature C and C++ environments exist on nearly every available platform. If one platform folds, much of your application should port to an new one trivially, and the rework will likely be in the user interface (a smaller task that dumping a .NET application and starting from scratch).
          • The argument that MS could possibly be a small fraction of the entity it is today -- well, it's just plain lame. People jumped on the Sun/Java bandwagon when Java seemed like nothing more than the pipe dream of a PhD locked in a corner office somewhere. There are also highly successful companies that continue to develop for Apple's platforms. Microsoft may be diminished in the future, but by the time it becomes a serious non-competitor in the marketplace, .NET will be a thing of the past and we'll all be programming in something totally different.

            And who said that just because you use .NET your application won't be portable. Unless you're using a framework like Qt, the GUI code has to be rewritten for each platform. As much as you'd like to think that C/C++ will just transfer from one environment to the other, most good apps require some sort of reworking to take advantage of OS-specific APIs for more than just GUI stuff. I've been doing x-platform development for a while now and C++ makes it easier, but it's not the panacea that you seem to think it is. .NET simply makes it easier to write the MS-specific portions of you app, and you can do it in C++. The rest of your code can be whatever you want -- you just may be stuck writing some hooks between your Win32/64 code and you .NET code, which is not very difficult if you know what you're doing.

            Come up with a different argument than "Microsoft might possibly vanish in a puff of smoke just like the Earth did in Odyssey 5". Or at least back up your argument with more than ".NET is simply the worst development tool out there." So far I see no reason why I shouldn't rely on MS's new framework for parts of my next app based on your statements above.
            • The argument that MS could possibly be a small fraction of the entity it is today -- well, it's just plain lame.

              It isn't "lame", because my argument is that using another company's success as the basis for your own success is just plain risky. Could Microsoft--or any company--be here today but not here tomorrow? Sure, recent history (Enron, Global Crossing, the dot-bombs, etc.) makes this clear. Microsoft is not immune to the discoveries of corruption (Enron) nor to the market forces (dot-bombs) that can make a company fall.

              As far as Java/J2EE goes, it is less risky than .NET, because Sun doesn't have total control over it. Sun's competitors are some of Sun's biggest licensing customers, and there are fewer single points of failure with Java than .NET.

              And who said that just because you use .NET your application won't be portable.

              Microsoft. It is extremely unlikely that .NET will be any different than prior Microsoft-imposed standards, unless there is a radical turnover of Microsoft's management.

              As much as you'd like to think that C/C++ will just transfer from one environment to the other, most good apps require some sort of reworking to take advantage of OS-specific APIs for more than just GUI stuff. I've been doing x-platform development for a while now and C++ makes it easier, but it's not the panacea that you seem to think it is.

              I never claimed C++ is a panacea but said that it can allow much less rework when used well (i.e., good C/C++ environments are common, good .NET environments are not). OS-specifics should be contained, so that the rework isn't such that it kills your product. Qt can be a great tool to this end.

              .NET simply makes it easier to write the MS-specific portions of you app.

              This could be one legitimate use of .NET, where it is peripheral to the rest of your application. Again, my point was about controlling risk. .NET is simply inappropriate as the "framework" of an application. It should not be the development tool.
    • Its funny you mention FUD. Its bad enough that companies like MS and Oracle and (*gasp!*) even Sun employ those tactics, but I suppose they do it out of some twisted duty to their stockholders. Whats your reasoning?

      .Net is no more spastic mess than Java 2/J2EE
  • This is the biggest disaster since the conception of Windows.
  • Yup, that's what M$ intended alright... contaminate as many programmers as possible by exposing them to the rotor libraries, so they can claim copyright infringement on anyone working on free implementations. not a bad plan, actually...
  • Does anyone know how I can benefit from this?

    Specifically I would like to offer customers solutions using Python. But if they want .NET integration, what do I tell them? It's not possible at all?

    I'd prefer not to lie to customers.

    If I could tell them that it's possible, at a price, that would be something! My customers would be prepared to pay! Not that they may ever need it once I get them their Zope/Python sites, mind you. I just want to leave them the option.

    Firstly, what would I have to do (pay) to get .NET commercially for Linux (Python)?

    Secondly, is it even possible to get .NET integration going in an effective manner? Are there tools/wrappers?
    • Does anyone know how I can benefit from this?

      With Microsoft Rotor, you will leverage your synergies to a higher ROI on your brain-share throughputs - now with new patented Microsoft "Shit-In-A-Box" Technology*!

      * Microsoft Rotor "Shit-In-A-Box" technology includes intellectual property licensed from Star Trek Ear-worm Developer Systems, Inc.

    • Specifically I would like to offer customers solutions using Python. But if they want .NET integration, what do I tell them? It's not possible at all?


      Perhaps if you can't bring .NET to Unix, you could bring Unix to .NET via Cygwin?

      You'd still be running a 'chalenged' server software, but it's an option.

    • You could use XML-RPC or somesuch to connect between Python and .NET code -- or wait for a .NET-based Python implementation (and, preferably, a good Linux .NET environment) and use your existing code then.
    • ActiveState is working on a python implementation for dotnet.
      http://www.activestate.com/Corporate/Init iatives/N ET/Research.html?_x=1
  • I don't know about you all, but I can't wait until Linux has the blue screen of death so I can see exactly why it crashed.
  • If I wanted to play with .NET, I would be running Windows right now. I'm not. I don't.

If you steal from one author it's plagiarism; if you steal from many it's research. -- Wilson Mizner

Working...