Slashdot Mirror


User: Kerg

Kerg's activity in the archive.

Stories
0
Comments
194
First seen
Last seen
Profile
(view on slashdot.org)

Comments · 194

  1. Re:Right on! on Miguel de Icaza & Nat Friedman On Mono · · Score: 1

    If they could port mercury to CLR, they can port anything.

    Then why does Eiffel# exist? Why isn't there a fully compliant Eiffel available on .NET?

    And why did they bother with Managed C++? You'd think if they can port any language to .NET they'd make damn sure it has a compliant C++ language. Yet they were unable to do it.

    Why is VB.NET so radically different from the earlier versions of VB? Why aren't the older VB versions supported on .NET?

    .NET does not have a universal virtual machine.

  2. Re:wake up and smell the standars on Miguel de Icaza & Nat Friedman On Mono · · Score: 1

    Yes, I've seen this.

    It only contains the specs for 70 classes. That is ridiculous. You cannot build an application with this.

    Where's the real stuff? There's nothing here that even resembles a GUI or database classes? There is a string, so I guess I could build a console app with it that outputs "Hello World".

    This is nothing compared to the thousands of classes Java provides.

  3. Re:wake up and smell the standars on Miguel de Icaza & Nat Friedman On Mono · · Score: 1

    I tried to find the ECMA specs from Microsoft web site on things like WebForms but couldn't. Maybe you could provide a link where they can be found?

    Also I'd be interested in taking a look at the ECMA specs on ADO.NET and ASP.NET. If you'd have links on those too, I'd appreciate it.

    Please stop posting things about things you know nothing about. Don't sprout rhetoric and stay to the facts.

    We're all here to learn. Perhaps we will once you show us where to find the full specs on all .NET libraries? (I'm really interested in WebForms).

  4. Re:Right on! on Miguel de Icaza & Nat Friedman On Mono · · Score: 1

    specs to ECMA so that any idiot can write a compiler to compile any language to the CLR

    It's not quite as easy as you make it sound. Consider for example ActiveState's Perl.NET implementation and the notes on their web site.

    In our Perl for .NET Research, we found that the initial approach of directly compiling to the .NET runtime had some limitations.

    [...]

    In the meantime, ActiveState will provide an alternate solution to make existing Perl code available from within the .NET Framework. Our PerlNET technology will allow .NET Framework code to access Perl code running outside .NET using the traditional Perl interpreter.


    It seems that the byte code of the .NET platform has been mainly designed to the C family of languages. Other languages will have difficulties to adapt. Even C++ has been turned into Managed C, and one of the main design goals of C# has been to offer the C++ developers a new language for Win32 development.

    more reliable for cross-platform stuff

    Care to inform us about those new reliability features?

  5. Re:Right on! on Miguel de Icaza & Nat Friedman On Mono · · Score: 1

    I don't think they would need to support C#. Just the .NET virtual machine.

  6. Re:Self interest clouds all else on Jepson Rebuts Petreley On The Dangers Of Mono · · Score: 1

    .NET is a brand new platform. It will run on many devices, not just Windows PC's. For example, it will run on X-Box's, and Hand Helds, and embedded devices.

    I doubt we will see .NET on embedded devices very soon. Right now Microsoft is pushing Java J2ME on WindowsCE to be able to compete.

    See Press Release here and The Register (ugh) article here.

    Microsoft's Stinger phone platform has failed to catch on, and Windows CE has so far been unable to dominate the hand held OS market. Microsoft has been forced to compete in this market, and in the embedded market with so many different platforms and hardware, you must be cross-platform, and you must provide an extensive cross platform programming libraries.

    This is why even Microsoft has been forced to use Java. There is no push seen here for .NET as there is really no evidence for .NET the platform being able to work in such environments. I doubt we will see .NET here very soon either (if ever). This is where Java (J2ME) platform has been really able to leverage its ability to work across so many many platforms.

    For now, .NET is just a Windows PC platform.

  7. Re:.net isn't that bad! on Petreley on Ximian and Mono · · Score: 1

    The reason for all the complicated adaptive optimization stuff is the bad design (performance-wise) of the bytecode.

    There must be more to adaptive optimization than that. After all, there are demonstrated cases where adaptive optimization does lead to a better performance than statically compiled "native" code.

    One such case I remember seeing (and I think was posted on Slashdot as well) was an algorithm that at one point grew it's datastructures big enough not to fit in the CPU's internal cache anymore (sorry, going by memory and too tired to find the site). Statically compiled code decreased performance-wise radically at this point. So did the HotSpot enabled Java code, but because of the run-time optimization, it was able to recover and ended up yielding better performance than the statically compiled (which was C or C++).

    You can even pre-compile your assemblies if you want to.

    Some people claim this is what gives C# better performance. I'm not convinced. Pre-compiling your assemblies would without a doubt make such adaptive optimizations of code impossible. I believe at least on server side apps which may be running extended periods of time, pre-compiling might be the less optimal solution. On the other hand, you can pre-compile Java code as well (see for example Tower Technologies) so precompiling is not an advantage C# has over Java.

  8. Re:The Real Question Here on Petreley on Ximian and Mono · · Score: 1

    In case .NET turns out to be a popular web services platform, should Linux hedge its bets against loosing server market share to MS by
    having an implementation of .Net available?"


    If .NET turns out to be a popular web services platform, then I guess there will be more web services around.

    Luckily, being able to put web services on the net doesn't require .NET. Linux can do web services already today.

  9. Re:Take a deep breath folks on Petreley on Ximian and Mono · · Score: 1


    With C# Microsoft are not sacrificing any performance, Java implementations are still processor and memory hogs.

    If both languages are compiled to byte code, and compiled at run-time, how does one not sacrifice any performance yet the other is a processor and memory hog?

    If you think that .NET is simply a language war

    No, it is a platform war. It's J2EE vs. NET. JVM vs. CLI.

    Not only that but Sun will sick lawyers onto folk who don't do it their way.

    Yes, I'm sure we can expect nothing like this from Microsoft. Have you checked their patents lately?

    It isn't just Microsoft that Gosling and co don't want to hear from, there are plenty of people arround who are serious heavyweights in the language design arena that got the brush off. Microsoft were very interested to hear my ideas, Gosling was not.

    Maybe Gosling didn't like your idea. So what? What do I care?

    "Heavyweight" companies like IBM, BEA, IONA, Apple, Novell et al seem to be able to work with Sun on the JCP. Claiming for example IBM has had no affect on Java and has been "brushed off" is ridiculous.

    C# has several very nice ideas about how to write programs that use XML.

    What features does the C# language have that makes it superior in handling XML document structures?

    I don't give a honk about write once, run anywhere. There are only two platforms that matter today, Linux and Windows. Running on Apple, Solaris, VMS, Irix, HPUX, Genera, BeOs, etc. is not something I am going to sacrifice 80%+ of my CPU for

    I'm glad you do have the luxury of not giving a honk about platforms such as HP-UX, Solaris, AIX or IBM mainframes.

    I, on the other hand, really can't go and tell all these companies who have made millions in investment to their Solaris, HP-UX or IBM hardware, to honk it. They need solutions, not useless rhetoric.

    or spend three times as long writing the program putting in optimizations that depend on the structure of the VM implementation or JIT.


    What kind of optimizations did you have in mind that depend on the JIT and take so much time to write?

  10. Re:.net isn't that bad! on Petreley on Ximian and Mono · · Score: 3

    you have a single API for all programming languages

    How is the API defined? Is there a interface definition language like CORBA IDL?

    Something that is still missing in the Java (even if they announced regex support for a future release

    JRE 1.4 does have a regex library. See here.

    It is definitely not more 'free' than the .net stuff (if MS really submits the specs to ECMA)

    It seems to me that it's just a different parts that are being 'free'. For some reason everyone thinks its a big deal to have the language and the virtual machine free. I on the other hand think it is far more important to have free implementation of the libraries that I use to build the applications.

    Java isn't part of ECMA or ISO standardization process. On the other hand, it was Microsoft that said ECMA is only useful for "rubberstamping" standards of a single company.

    Java does have JCP which seems to work rather well. Open Source groups such as Apache, Enhydra and JBoss are represented. All libraries are extremely well documented (they have to be in order to have implementations on all platforms, unless they can be written in 100% Java), and Open Source implementations of those libraries do exist.

    Microsoft has promised to put C# and the virtual machine through the ECMA process. However, if most C# developers come from the Win32 background and heavily use libraries such as ASP.net, WebForms, ADO.net then how will those applications be portable? If you look at the struggles Wine has gone through, it is obvious porting the libraries is far from being trivial. And these libraries have not been submitted to ECMA.

    I have never had a real need to have the Java language or the JVM standardized by ISO or ECMA. If someone wants to make additions to the Java language, they can create their own language, and call it Pizza. They may use the JVM or not, up to them. I've never had a need to go change the JVM implementation either. They have been very stable.

    The libraries however are a different story. Lots of bugs, lots of things I want to change. And it's quite plausible to do so. There are Open Source implementations written for Servlet, JSP, EJB, JDBC, JTA, JavaMail, JDO, etc etc API's. This is far more important to me as an application developer. It directly affects me, where as Java the language or the virtual machine being free has far less effect (because they do work, and do work quite well). Not to mention as an application developer I'm no expert in designing languages, let alone understanding the details of technology like HotSpot.

    So even still, I believe if you need to write cross platform applications (within the next 3-4 years, and more complicated than Hello World), I think .NET is not an option.

    The JavaVM is quite limited to Java as language. Yes, there are other languages for the JavaVM, but the VM isnt really flexible enough and you dont have things like pointers available which you need for system programming.


    You said you rarely need to use pointers yourself. I agree. And when I do, I don't want to do it inside the virtual machine and compromise it's stability or security. For any system programming I will use C. There are plenty of ways to have a C program and a Java virtual machine interoperate. I don't want that interoperability to crash the virtual machine.

    Java's performance sucks because of several design mistakes

    Adaptive optimization of HotSpot really does make a difference in many cases. I haven't seen something like that mentioned for the CLR (is that the correct acronym?)

  11. It's not odd. on Diablo II: Knickknacks Nicked · · Score: 1

    Is it just me, or is it a bit odd to be reporting on the disappearance of items that never existed in the first place?

    I don't think it is any more odd than complaining about missing email from your inbox. It could be equally well said that all that mail never existed in the first place.

    Suddenly it's no longer just 0's and 1's anymore.

  12. Re:Wasn't this expected ? on Mono Unimplementable? · · Score: 2

    However, with .NET it all comes in one attractive package all ready to go

    Ready to go and do what?

    I mean, if you look at IBM's Java offering for example, they heavily tie their VisualAge IDE to their WebSphere application server and other tools. That's a ready to go package to do pretty much everything you need, web apps, middle tier, web services, EJB, JDBC, whatever. All bundled with an IDE with version control, debuggers, help blah blah blah.

    The matter of fact is most people tend to not want this kind of tie up. Some people like it, and I bet they're all happy IBM customers and more power to them, but I for example would not even want a package like this for my Java development.

    A ready to go packages are available for Java. They even let you pay lost of $$$ for it. I personally like it much better where I can choose the different components for development myself.

    And as you point out in your post .NET will even talk to legacy Java code via SOAP

    And Java uses SOAP to talk to legacy Windows platform with apps written in C#, VB, whatever.

    That's what makes .NET so great. It's got all of the advantages of the Java platform, without the downside of only being able to use the Java language.

    I'll buy this if I ever see a cross-platform ADO+, ASP+ or COM+. Until then, Java is much much more attractive solution to me. Not to mention it is much more mature platform.

    It's also worth noting that Java isn't nearly as cross platform as Sun would have you believe.

    It's good enough for me. I've written apps that people deployed on Windows (NT and 2000), Solaris, Linux, HP-UX, MacOS and possibly others. I can tell you I don't know much anything about the Unix/Linux platforms. But my apps work in those environments and that I am happy about. As far as I'm concerned, Sun made good on their promise of cross-platform and WORA.

    A Free Software implementation of .NET would solve that problem nicely, and it would quickly become a very cool development platform.

    But there wont ever be a full blown .NET platform under a Free Software license. The only parts Microsoft has made available is the C# language and the virtual machine.

    That's a far cry from the .NET platform.

  13. Re:Wasn't this expected ? on Mono Unimplementable? · · Score: 2

    but in my testing with C# - it's pretty damn fast.

    Micro benchmarks mean very little. Some people claim they've seen 50% speed improvement in C#, some say they see Java performing alot better. There are some legit arguments that can be made about C# approach to compiling at install time which will decrease the performance in comparison to an adaptively optimized (HotSpot) Java.

    My personal guess is they're pretty close to one another. There's no single technical issue on either side that would make one faster than the other.

    I'd be interested to see any legit benchmarking though.

    unpredictable in terms of platform compatability, speed, and version

    What makes C# or .NET any different in this regard? I've been extremely happy with Java's compatibility btw. I know exactly what I can expect with each iteration of the SDK.

    A good GPL'ed C# compiler on Linux would win me over.

    Not me. Java the language matters very little. The language can't do squat. The same applies to C#. It's the libraries that count, having JDBC, RMI, JNDI, EJB, Servlets/JSP is what drives me to use Java. What language they operate on makes no difference. Having a huge, cross platform libraries is a blessing.

    Having a C# compiler in GPL means squat. If you get ADO+, COM+, ASP+ then you may have a case. I doubt Microsoft will let you have all that on Linux though.

    GPL Java JIT that's good for a while to come

    I don't need one. I'm not interested in writing compilers. I'm interested in writing other applications though.

  14. Re:Another reason... on Mono Unimplementable? · · Score: 1

    We (as a community) have absolutely no control over its cross-platform issues or anything else to do with it

    Yes, you should also remind the Apache folks, JBoss people, and the Enhydra group that despite them being part of the Java Community Process they're just puppets and have no way of influencing anything that has to do with Java.

  15. Re:Wasn't this expected ? on Mono Unimplementable? · · Score: 1

    On top of that they are adding a powerful easy to use RPC system based on SOAP

    SOAP works perfectly well with Java too. This is not an advantage .NET has over J2EE.

    has really come up with something in their .NET architecture

    People keep repeating this but rarely come up with any technical arguments for what it is that makes .NET so great? What can it do that I cannot do in Java?

  16. Re:Wasn't this expected ? on Mono Unimplementable? · · Score: 1
    The competing architecure in Java is J2EE. There are already several Open Source implementations out, that implement either the whole platform, parts of it, or the new web hype services.

    JBoss Enhydra Jakarta Jetty jUDDI pUDDIng Resin

    Just pick a project and start contributing ;-)

  17. Re:Web Services using XML vastly superior to Java on Microsoft To Assist Ximian In Producing Mono · · Score: 1

    Why would you do that? Just to spite Microsoft?

    It has nothing to do with spiting Microsoft, or anyone else. Why does an admin decide to block any of the other RPC mechanisms? To secure their system. It's not like its a huge task for an admin to let a protocol like IIOP through their firewall, yet they decide not to. SOAP is no differnet there, it's just an RPC mechanism. If the web service itself is not secured inside a well defined sand box, it is a security risk to have SOAP invocations being piggybacked on HTTP to the service. Either you know all of the services running on your system, and trust them to be implemented securely, or block SOAP invocations.

    VB is still the most popular programming language out there, for better or for worse.

    There's been a huge drift from VB to Java in the last couple of years. This what has Microsoft worried, they're losing their developers. I've personally helped several VB developers to migrate to Java, and I know it is a common occurence in many businesses.

    VB is lacking behind Java in terms of programmer productivity, especially in environments such as the .NET and J2EE platforms target. C# alleviates the problem, but it's a whole new language that the VB developer needs to learn. VB.NET is not a trivial upgrade either. Many VB develoeprs have said the move to VB.NET requires quite a bit of a learning effort.

    J2EE has a two year head start on the application server market. The major J2EE platform vendors already support SOAP, UDDI, WSDL, etc. There's a definitive movement amongst developers from VB to Java. No wonder Microsoft is worried.

    Duh!! How are they "not impressive"? If you knew even remotely anything about them you would know that SOAP in particular is only as smart as its payload - its a fricking RPC packet wrapper.

    Yes. Your point was?

    XML-RPC is nothing new. SOAP standardizes it. That's good. I'm glad IBM dragged Microsoft through the process. But I don't find the spec particularly impressive, from a technological point of view.

    UDDI, the same thing. Put together quite quickly by Microsoft, IBM, and some other companies (Rosetta?). It's now borrowing features from ebXML registry which went further in their definition of business registries. We shall see what comes out of it (will they merge?).

    WSDL, it just defines the contract of the web service. Compare to IDL. What's impressive here?

    go read some of the white papers and come back with an opinion.

    I have. I wonder whether you have any technical arguments to present to the discussion, or will you just continue with your trolling.

  18. Re:Web Services using XML vastly superior to Java on Microsoft To Assist Ximian In Producing Mono · · Score: 1

    So your point is .NET is impressive because it contains some open standards, it uses a slow HTTP protocol as a transport for RPC (mostly to get around firewalls, I suppose, which the net admins will block as soon as they figure out what is going on... hmm wonder who will be the first to implement a firewall that automatically drops all HTTP requests with a <SOAP> tag in the payload...), because MS has a monopoly and Java has failed (which it hasn't).

    I'm sorry but you fail to impress me.

    Web Services are one thing, its a nice idea for integrating systems that don't need tight coupling. For anything else, it blows. Web Services are only a part of the .NET platform. Yet you did not come up with any other explanations why .NET is such an impressive technology.

    Java most certainly has WORA. Without Java I would not ever bother with Linux/Unix implementations of my software. Now I can with very little effort. Java has not failed in the market place. If you actually had had a look at the market place (the same one .NET is targeting and competing with J2EE) you would not make a statement like this. The whole point of .NET (formerly known as DNA) is to compete with Java because Sun is winning A) the application server market place B) the developer mind share.

    I have nothing against SOAP, UDDI, or WSDL. But they're hardly impressive. Yet another spec in the long long long line of XML-this and XML-that. Who is impressed by that anymore?!

  19. Re:Ximian, don't be silly. on Microsoft To Assist Ximian In Producing Mono · · Score: 1

    Still, Sun could give a lot more support to the Free Java implementations, so they aren't playing a constant game of catchup to a moving specification.

    Again, which specification are you referring to, the JVM?

    I don't have the impression that the virtual machine specification is a constantly moving target, quite the opposite. Sun is often critized by their unwillingness to do *any* changes to the JVM spec. So what is the catchup the Free implementations are forced into?

  20. Re:Ximian, don't be silly. on Microsoft To Assist Ximian In Producing Mono · · Score: 1

    I do not know how it can be said that Java competes with .NET. "Competes" implies mutual exclusion, and I believe .NET-enabled applications are allowed to use Java's rich APIs.


    I was referring to the J2EE platform there actually... J2EE competes with .NET, as it did compete with MS DNA, which is what the .NET was formerly known (minus stuff like Passport, etc).

    However, most of the arguments are like saying that CORBA competes with Java, because C++ programs can then access Java APIs.


    I think its a platform issue... you code to the .NET platform or you code to the J2EE platform. Java or C# (or C++) are just languages you use, it is whether you deploy your application on the .NET, J2EE, or CORBA platform.

    There's no money in programming languages for any company, not Sun, Microsoft or IBM. However, there is a multibillion dollar market in the platform, and companies such as Microsoft, BEA, IBM, Sun and Oracle certainly want you to deploy on theirs (and whatever specification their platform may implement). That's where the competition is, and J2EE directly competes with .NET there.

  21. Re:Ximian, don't be silly. on Microsoft To Assist Ximian In Producing Mono · · Score: 1

    It's scary that Microsoft has made C#/.NET more accessible to Free software development than Sun has made Java/EJB.

    How is .NET more accessible than Java/EJB to Free software development?

    I'm doing Free software development on Java/EJB and I'd really like to know what the difference is to .NET platform?

  22. Re:Where are MS rivals on this? on Microsoft To Assist Ximian In Producing Mono · · Score: 2

    I don't know about IBM but I doubt Sun feels fucked by the OSS community. There are several Open Source implementations of the J2EE platform or parts of it available and being developed. It's mostly that the Linux community is somewhat ignorant of them. Plus, of course they have more than 20 commercial companies developing implementations the J2EE platform, including IBM, which guarantees competition in the platform implementation and I doubt we will see in the same scale with the Microsoft platform.

    For Open Source J2EE, check the following:
    Jakarta
    JBoss
    Enhydra
    Jetty
    Resin

  23. Re:Ximian, don't be silly. on Microsoft To Assist Ximian In Producing Mono · · Score: 1

    And why should anyone ally with Sun and Java? I've read their license and billj's justification for it. Why should anyone surrender their rights in this manner?

    What does writing Java software or implementing Java API's have to do with Sun's community license?

    There are plenty of Java implementations out there, both commercial and open source, that directly compete with .NET. Yet you never hear any of them mentioned in /. news. Seems the editors are to keen on spreading the .NET hype.

    Jakarta
    JBoss
    Enhydra
    BEA Weblogic
    IBM WebSphere

  24. Re:It's my body, I'll poison it if I want to on The Glories of Red Bull · · Score: 1

    It's not the caffeine that is the problem, its the taurine. Try to follow the discussion.

  25. Re:It's my body, I'll poison it if I want to on The Glories of Red Bull · · Score: 1

    You know what it is, and you should understand what you're getting into when you drink it!

    Can kids buy this stuff?