Mono Poises to Take Over the Linux Desktop

Edd Dumbill writes "Miguel de Icaza and the Mono team recently hosted a two day open meeting in Boston. O'Reilly have just published my report of the meeting. Highlights include Miguel's view that 'C is dead!' and the Mono approach to dealing with Microsoft patents on .NET."

Comments

C is Dying?

[Anonymous Coward]

C is dead? Has Netcraft confirmed this?

Re:C is Dying?

[Anonymous Coward]

Maybe because the D programming language is set to replace it. As if we'd be so lucky.

Re:C is Dying?

[Anonymous Coward]

(!C) ? (alive = 0) : (alive--);

Re:C is Dying?

[Canadian1729]

Saying C is dead is such a stupid comment, it invalidates anything else he might say.

Re:C is Dying?

[pavon]

In a recent interview, Netcraft quoted C as saying, "I feel happy" and "I think I'll go for a walk now". However Miguel was quick to point out that "C is not fooling anyone" and "is just being a big baby." Unfortunatly, the interview was cut short, and Netcraft will have to wait until next Thursday to confirm the C's death.

Re:C is Dying?

[Ironsides]

Obviously this guy has never been in an Engineering College lately. We are still learning C. Not C++, Not C#, Not Java, C. As in where you have to declare all your variables at the very beginning and can't declare them anywhere else. It's how we program microprocessors (among other things) in something other than assembly. And I highly doubt they will make a Java compiler for PICs, Motorolla HC11 and HC12's and various other devices any time soon. Besides, I here places are still looking for COBOL and FORTRAN programmers. Seeing as how those two are still here, C still has a long way to go before the white lillies come around.

Re:C is Dying?

[dglo]

Maybe because the D programming language is set to replace it.

C is the successor of B, which was a rewrite of BCPL. Therefore, the only true successor to C would be P (the next letter after B and C)

Wayne's World

[wtlssndlssfthlss]

"I once thought I had Mono for an entire year, just turned out I was really bored..."

c is dead...

[dieyack]

long live c!

Re:c is dead...

[gstoddart]

Indeed. Since most damned operating systems are written in C I don't think we'll exactly see it go away anytime soon.

Operating systems is where C comes from, and vice versa.

(Yeah, whatever, I'm a C geek. To me, and array of pointers to functions returning pointers to arrays of characters seems like a damned fine idea! =)

Back in my day, we jusy wrote straight to the registers on the device and we were glad! Damned punk kids.

.NET

[Anonymous Coward]

I realize that this is an unpopular opinion here on Slashdot, but C# is actually a pretty cool language and the .NET runtime is a promising platform. Microsoft didn't just dream this up overnight... they had a lot of smart people working a long time creating this beast.

It would certainly benefit us to learn about these technologies and leverage them, rather than to unilaterally declare them evil, wrong, stupid, etc. and just bury our heads in the sand and pretend they dont exist.

Re:.NET

[ivern76]

I disagree. After a few months of working with C#, I have come to the conclusion that it's nothing but a bastardized clone of Java that had every little piece of sugar they could snarf from VB and C++ added to it, whether it made sense or not.

For example, the braindamaged distinction between structs and classes as value and reference types respectively. What if you want to make both value and reference objects of the same class? Or treat the same object as value or reference as needed? The stupidity becomes even more apparent when you find out that the System.Array.Initialize() function will only initialize value arrays, not reference arrays (WTF? Is there ANY reason to even have the [,] array creation operator if you still have to construct the members one by one after creation?). I could go on and on about "features" that were clearly hacked on in five minutes to fit some deadline, with little thought or care.

C# does have some nice things that, say, Java lacks. Operator overloading, automatic boxing and unboxing of primitive types, and properties come to mind. The first two, AFAIK, will be in Java soon, and properties are just syntactic sugar to replace observer methods. Honestly, I'm not impressed, and I don't intend to use C# ever again unless I absolutely have to.

Re:.NET

[Rascasse]

.NET is good because it takes the best from languages that are already in existence. It's not like there is anything revolutionary in C# that isn't in any other language out there. I used to be stuck in this PC-centric view of the world. Imagine my surprise when I bought a Mac and realized that NeXT had fantastic things like Internet-enabled Distributed Objects available long before much of the world even knew what a web browser was. Yes, C# is good. But the only people that I know that have been blown away by it are those that didn't stray far from Microsoft solutions and were never exposed to tools available from other vendors. Welcome to what much of the rest of the world had available to them in the form of Java since the late 1990s. And before the fanboys come out - yes I know that C# actually improves on some of Java's deficiencies. But I do take issue with the assertion that C# was made possible thanks only in part to a concerted MS R&D effort. It wasn't.

Re:.NET

[willdenniss]

here here.

And, unlike .NET - Java is fully featured across all of it's supported OS's.

Will.

Re:.NET

[jfengel]

Autoboxing yes, but I don't believe that they have operator overloading in mind. I've heard Gosling say that it adds a lot of complexity to interpreting a program while adding relatively little expressive power.

(A picture of Gosling standing in front of a slide explaining what's wrong with operator overloading).

I'm definitely looking forward to autoboxing and the new printf-style methods. (Maybe variable-argument methods were what you had in mind?)

Re:.NET

[grennis]

What if you want to make both value and reference objects of the same class?

Value/Reference is a class-based distinction. It has nothing to do with instance, and I can't say I've ever had a real world problem with this. Can you?

Or treat the same object as value or reference as needed?

Real world example?

the System.Array.Initialize() function will only initialize value arrays, not reference arrays

Of course. C++ is no different. Remember how in C++ you have to write "delete [] x" instead of "delete x"? Simply because you have to communicate to the compiler that it must call the destructor on each array member... but the constructor is much more difficult. What if the array element type has no default constructor?

I could go on and on about "features"

And I could equally as well refute your points.

It is ignorant to simply dismiss .NET or C# because it is something you don't know and don't understand. Take the time to learn it... there are valid criticisms, but you haven't found them yet.

Re:.NET

[nemiak]

" Microsoft didn't just dream this up overnight..."

Like most/all of Microsoft's "innovations", .NET was *purchased*, not created, by Microsoft.

Microsoft inherited what become .NET when they purchased Colusa Software on on March 12, 1996.
At this time C/C++ and VB environments already existed for Colusa's "OmniVM" (which became the .NET CLR?).

To clarify - I do alot of work with C# and find it to be the least unpleasant Microsoft development environment I have experienced.

Score 5, Troll?

Re:.NET

[SensitiveMale]

Actually what he should have typed

"I realize that this is an unpopular opinion here on Slashdot, but C# is actually a pretty cool language and the .NET runtime is a promising platform. Microsoft didn't just dream this up overnight... Sun had a lot of smart people working a long time creating this beast."

Re:.NET

[Anonymous Coward]

man, sometimes i don't even know why i look at the discussions of programing and scripting languages. I hardly ever know what people are talking about.

C# This is essentially Microsoft's new version of the C programing language. Java Sun Microsystem's cross-platform programming language. VB Visual Basic. Microsoft's language for Rapid Application Development. C++ Popular extension to C which adds support for Object Oriented Programming. Struct A concept used to group variables together. You can refer to the whole group of variables using a single name. Class A struct on steroids. Not only does it group variables together, but it also groups functionality with the variables. Reference type A variable that points to the real data. You can have many reference variables that all point to the same data. Value type A variable that is the real data. Boxing The process of taking a reference type and converting it to a value type. Unboxing The process of taking a value type and converting it to a reference type. Operator overloading Redefining what an operator does. For example, if the equal sign in the following code assigns the value 2 to the varialbe a,

a = 2;

Operator overloading could redefine the equal sign to mean something else (such as compare if a equals 2).

Property You can think of this as a wrapper around a variable. When the variable is accessed, you can do some automated processing before returning (or storing) the value. In code, you can treat a property just like a variable.

Re:.NET

[Bingo Foo]

No, .NET is full featured across all one supported OS.

Re:.NET

[atlasheavy]

actually, .net is fully featured across all of its supported OS's as well. It's just that its supported operating systems are Win98, WinME, Win2k, WinXP, and Win2k3. And Longhorn.

Re:.NET

[IllForgetMyNickSoonA]

Yeah, sure. Last time I checked, there was no operator overloading in Fortran either. Nevertheless, it is still the preferred language for mathematicians (among others).

Operator overloading, while in some occasions quite handy, must be used with great care. Unfortunatelly, for quite a lot of programmers out there, it seems to be misused as a yet another way to obfuscate a program into oblivion! If you ever get to analyse the code written by a BOOZFH (Bastard Operator Overloading Zealot From Hell), you'll know what I'm talking about.

Re:.NET

[ghamerly]

I believe you have boxing & unboxing reversed. I didn't know the definitions before this, but the ones you gave didn't seem intuitive to me, so I searched google for "autoboxing", and found several definitions that are the reverse of yours.

Re:.NET

[ivern76]

What if you want to make both value and reference objects of the same class?

Value/Reference is a class-based distinction. It has nothing to do with instance, and I can't say I've ever had a real world problem with this. Can you?

Or treat the same object as value or reference as needed?

Real world example?

Easy. Value types allow you to transparently create copies of data, essentially letting you deal with immutable objects without the added complexity of calling a clone() operation every time you need a new copy to work with. This has nothing to do with class characteristics, and everything to do with the particular code you're using.

the System.Array.Initialize() function will only initialize value arrays, not reference arrays Of course. C++ is no different. Remember how in C++ you have to write "delete [] x" instead of "delete x"? Simply because you have to communicate to the compiler that it must call the destructor on each array member... but the constructor is much more difficult. What if the array element type has no default constructor?

If it has no default constructor, then you make it an error for value types (duh) and a nop on reference types. Alternatively, you could make it an error on both. If there's no default constructor, you don't call Initialize(). Your comparison to C++ loses, because 1) we're not discussing C++, and 2) we're not discussing "telling" the compiler how to do a job, we're discussing how the compiler doesn't do said job no matter how you "tell" it to.

I could go on and on about "features"

And I could equally as well refute your points.

That's great to hear. Start any time you want.

It is ignorant to simply dismiss .NET or C# because it is something you don't know and don't understand. Take the time to learn it... there are valid criticisms, but you haven't found them yet.

I wouldn't go as far as to call you ignorant, but I know for certain it's stupid to dismiss my claim without actually addressing it. If you scroll up and actually read my post and yours, you'll notice that you either didn't read my claims properly, didn't think the matter through properly, or didn't type your reply properly.

Why do I claim this? Well, 1) you don't seem to understand how you can use value and reference types interchangeably in a real world situation, yet you go ahead and talk about it. 2) You try to address my issue with Initialize() by pulling a textbook Chewbacca defense with C++'s delete[] syntax. And 3) you tell me there's "valid criticisms", yet you fail to mention any. All you know is mine don't work, yet you also fail to adequately explain why.

In the end, I have no option but to classify you as a troll until you prove otherwise. I suggest you start doing so by addressing my initial post with clearly thought out answers.

Foot-in-Mouth Disease

[adun]

Aside from being a the primary source of Mediterranean winds, Icaza has apparently forgotten about that whole "Linux" thing that is built on that whole "UNIX" thing that was built using that whole "C" thing. I applaud his salesmanship. I deplore his view that the desktop is equivalent to the operating system.

Re:Foot-in-Mouth Disease

[milest]

The article talks about C dying for *user* applications explicitly:
"A great deal of serious end-user application coding on Linux still goes on in C or C++"
"Where does this place the future of the Linux desktop..."
"The expectation for GNOME 3.0, however, is that a lot of the platform will use Mono, rather than the C implementation it has now"

Probably no one writes kernels in C# and there is less and less reason to write user applications in portable assembly. Programming languages are tools, and we use different tools for different jobs. I prefer to use higher level languages where it is feasible because I'm more productive. I use C where I have to, because it has features that other tools don't.

C is not dead

[Yaa 101]

You need a solid underground when programming...

MS and solid? hmm...

WTF?!?

[Anonymous Coward]

Will Mono Become the Preferred Platform for Linux Development?
by Edd Dumbill

Why the hell would I believe someone with the name of Edd Dumbill?

That must be the "Alan Smithee" of the so-called "tech writers." It's probably just dumbshit Alan Seeburgh in disguise, playing hooke from CNN.

Um, no.

[ivern76]

This falls under the "I can't believe what I'm hearing" category...Mono is *not* ready as a plug in replacement for .NET, and it won't catch up before MS releases 1.2...for the foreseeable future, it's trailing behind the Windows implementation and is not likely to catch up.

I see PyGTK as a much more reasonable (and WORKING) alternative to C programming for people who want to write Gnome apps. Or GTK--, for that matter. Mono currently has crappy System.Windows.Forms support (even with Gnome#), broken serialization support, the list goes on and on.

Re:Um, no.

[IAmTheDave]

Yeah, it's crappy. But heck, it's a move in the right direction. I know certain people (cough... slashdotters) are quite protective of the purity of writing unmanaged code.

But that said, I have never developed software more rapidly than in C#. .NET has trippled my productivity (on the Windows platform) and my approval rating at work has skyrocketed as I have rolled out several solutions on .NET that are stable, solid, and effective.

I really wish that Linux had a stable .NET equivelent (sp?) - cause then I could completely abandon Windows at home. But as long as I can code C# with such RAD success, Linux is going to have catching up to do.

I applaud MONO. It's far from ready, but keep going. Keep working. Please - make MONO (or something else) as good as C#, .NET, and Windows Forms (and hell, ASP.NET) is. RAD is so important in the real world, that Linux would only benefit - incredibly - from a .NET like solution.

Re:Um, no.

[miguel]

System.Windows.Forms is not part of the Mono 1.0
release as you well point out.

But Gtk# 1.0 is part of the Mono 1.0 release, and
unlike Windows.Forms it is very mature and stable.

Love,
Miguel

Re:Um, no.

[cerberusss]

I really wish that Linux had a stable .NET equivelent (sp?) - cause then I could completely abandon Windows at home.

Why don't you use Java?

Re:Um, no.

[Brandybuck]

.NET has trippled my productivity (on the Windows platform)

Anything other than MFC would triple your productivity.

Re:Um, no.

[bangular]

There isn't a demand for anything like that on linux though. Linux programmers tend to know a few different languages (a couple of high level perl/python like ones, C or C++, bash/tcsh/ksh/etc shell scripting, and sometimes Java) and use what ever one is appropiate for the job (I over generalized, I should say, that's what I do). You know pretty early on what language your program needs to be written in. The problem with just using C# is your still just using one language. You're not using the right tool all the time. It's the same reason I tend to disagree with Sun's attempt for Java to take over the world. You end up driving in a screw with a hammer.

MonoDevelop IDE

[gmajor]

Doesn't the Monodevelop IDE look suspiciously like Eclipse?

Re:MonoDevelop IDE

[k_head]

Why develop your own IDE? Why not write a plug in for eclipse?

Bah, you probably wanted something written in C#. So much wasted effort.

Linux contracts mono?

[Black Art]

Are we supposed to clap?

Re:Linux contracts mono?

[mesach]

no no... its mono, not the clap

Learning and applying.

[Faust7]

It would certainly benefit us to learn about these technologies and leverage them, rather than to unilaterally declare them evil, wrong, stupid, etc. and just bury our heads in the sand and pretend they dont exist.

I think that about a lot of Microsoft technologies -- but to be fair, I'd say that the mere existence of Mono is evidence that a process such as you have described is already in motion.

C lives

[Anonymous Coward]

There will still be times for the necessity of optimizing for something for which the .NET JIT compiler will not be sufficient.

Also, .NET is much like Java, and C hasn't died from the adoption of Java.

Exploits

[bored1]

Great, now users will be stuck fixing Windows and Linux exploits while using Mono

Overnight

[SuperKendall]

I realize that this is an unpopular opinion here on Slashdot, but C# is actually a pretty cool language and the .NET runtime is a promising platform. Microsoft didn't just dream this up overnight...

Yes, it took at a least a couple of days to copy all the Java libraries and ReCapitalizeMethods,

Seriously though, .Net is a nice language with some advancements over Java, but not different enough from Java to make its existence worthwhile. It's just leading to a lot of duplication of effort across the world (like Ant and Nant, or JUnit and NUnit).

Now if they'd come up with something like Haskel# as the primary language (instead of hamstringing other languages and making all of the core libraries Java like) or something really different that actually advanced the field of programming as a whole, then I might be more appreciative But as it is I see the tremendous duplication of effort across the world to do the same things in Java and C#, and it just makes me sad.

Re:Overnight

[miguel]

You might want to look at Nemerle (www.nemerle.org)
a nice functional language that runs on .NET and
Mono.

Re:Overnight

[primus_sucks]

But as it is I see the tremendous duplication of effort across the world to do the same things in Java and C#, and it just makes me sad.

I think this problem would be solved by writing things in Scala. Scala compiles to .Net or Java runtime and is a far more advanced language than C# or Java.

Re:Overnight

[gabebear]

While I agree with you on every point, I don't see any of those as being very big problems.

I still give Java the upper hand overall because it's licence. Unless Microsoft blesses Mono with a licence to use the full API, I can't see using it. Mono keeping the APIs they may not have a licence for in a seperate code-base is not enough for me.

The only modern language that is wide-spread and licenced sanely is Java.

Re:Overnight

[ajagci]

Seriously though, .Net is a nice language with some advancements over Java, but not different enough from Java to make its existence worthwhile.

Maybe not if your idea of programming is limited to hacking J2EE server-side components and a few flaky GUI apps.

For the rest of us, C#'s better native interface, value classes, operator overloading, fixed generics, and fixed exception handling alone are worth it as a language alone.

Add to that that there is a high-quality open source implementation (Mono) and a complete stack of non-proprietary open-source libraries (Gtk#, etc.), and people have really strong reasons for preferring Mono to Java.

But as it is I see the tremendous duplication of effort across the world to do the same things in Java and C#, and it just makes me sad.

Well, do you think that Microsoft and the OSS community are spending time on .NET and Mono respectively for the fun of it? Sun forced this duplication: through their proprietary license requirements for Java and through their poor handling of Java's technical evolution. In the end, Java's evolution is primarily driven by what makes money for Sun, and that's why Sun has neglected client-side Java and Java numerics, for example. And the reason why they could get away with that is because they have retained legal control of Java.

So, if this "duplication of effort" makes you "sad", complain to McNealy and Gosling and Schwartz and all of the other Sun managers: it's their responsibility, it's their greed, and it's their fault that it has come to this.

You can't beat Microsoft....

[fatboy]

.....at their own game.

You can't a better .NET than .NET

Anyone remember "a better Windows" than Windows called OS/2?

Renovate or Innovate?

[capz loc]

One of the problems I have with Mono is that it reimplements Microsoft's .NET framework instead of relying on its own. I understand that it is a lot of work to create such an integrated development platform from scratch, but I feel that the benefits of having a truly open/free/whatever solution will far outweigh the extra time that it would take. I am not a programmer, so I don't know much about the intricacies of .NET, but I can't possibly imagine that it is perfect in every way and that nobody can think of things that could be added/changed in it.

What Linux desktop?

[Operating Thetan]

It seems we get 5 articles a week explaining how Linux isn't ready for the desktop. How can Mono take over something that experts assure us doesn't exist?

What language is .NET written in ??

[konmaskisin]

Languages written in C:

perl
python
ruby ....

If I wanted MS .Net, I'd run MS...

[Spicerun]

As it is, I don't want to even attempt to emulate another 'grand MS idea'...especially since there are already superior non-MS systems out there that puts .net to shame. No, I'm not going to cite those systems...do your own research. You'd be surprised.

Die Mono Die!

Platform Independence

[WombatControl]

I've been playing around with the Mono implementation of C#, and it's pretty good. It's not quite as good at RAD tasks as Python, but it has some advantages, and the syntax is much easier to play with than C or Java (but again, not quite as easy as Python, but I'm biased in that regard).

However, Mono suffers from the fact that they're trying to play follow the leader by following Microsoft's implementation rather than creating a system of libraries from scratch. Microsoft has a history of pulling the old "embrace and extend" trick, and I fear something similar may happen here.

My guess is that Microsoft will significantly alter the .NET APIs for Longhorn, leaving Mono behind with older legacy libraries that are no longer interoperable with the Microsoft compiler and the rest of the Windows-using world. Needless to say, that would be bad for the Mono team.

Still, if Mono can remain independent, it could very well have a bright future. The Mono team has done a great job of implementing most of the 1.0 .NET API, and the mcs compiler is pretty fast. The GTK bindings are quite nice for such an early release.

Still, the cognitive dissonance of compiling a Linux program and getting a file with an .exe extension is rather difficult...

Re:Platform Independence

[WombatControl]

Uhm, why would MS do this? .NET is designed to be and sold as a "cross platform" solution--a real way to do Java's "write once, run anywhere" line.

For the same reason they did it with Java - if it's "write once, run anywhere", then why would you buy Windows licenses? Microsoft (quite naturally) wants everyone to run a Windows server and a Windows client, and having Linux be able to take either role with ease doesn't give them the leverage they need to continue their marketshare.

Re:Platform Independence

[Brandybuck]

.NET is designed to be and sold as a "cross platform" solution

Don't bogart that joint, my friend. It sounds like good shit!

Miguel is dead!

[Nailer]

- C.

Re:Miguel is dead!

[epiphani]

Slightly offtopic, but a great quote about C...

C is a language with the flexibility and speed of assembly combined with the ease of use of assembly.

Re:Miguel is dead!

[orthogonal]

C is a language with the flexibility and speed of assembly combined with the ease of use of assembly.

Of course, that's tongue in cheek, and yes, C's concept of arrays is perhaps too tied to a specific kind of machine architecture (and to BCPL's concept of arrays),

but

while C is far more portable and easier to code in than assembly, one great feature of C is that, if you know a particular machine's assembly language, you can look at your C code and pretty easily guess what assembly instructions will be generated from it.

I recall when I was first learning C++ -- and no, I'm not a C++-hating C bigot -- one thing that annoyed me was the realization that what seemed a straightforward C++ function call -- or even more straightforward, an operation like i + j, when performed on class instances, could be much much more involved.

(Consider the example I just gave, i + j: in C++ this could easily be i.operator+( j ), which might involve an implicit converting function or constructor to make j into the right type for i's operator+, and operator plus could easily be a virtual function that did some work and then called a super-function to do the rest of the work, and it wouldn't be a good operator+ unless it returned by value, which would mean that there'd be a (probably implicit, in the code) call to the constructor for whatever class i was an instance of. And we haven't even mentioned Handle-Body Idiom or Template Method Pattern or Pimpl Idiom or exception handling yet.)

I soon got over my annoyance when I realized that it was precisely this ability to concentrate this much -- and much of it mostly automatic, as with converting functions of constructor calls -- into the semantics of addition, i + j, that makes C++ such a powerful tool in the hands of the right programmer.

The trade-off, of course, is not being able to see the assembly language instructions in your mind anymore. The good programmer chooses his tools and his trade-offs.

So to bring this back onto the topic of the post, I highly doubt C is dead or even dying or even a bit ruffled around the edges -- there are still, and for the foreseeable future, there will be, very very good reasons to use C -- it's the portable lingua franca of programming, and it's a remarkably clear language that allows one to concisely describe what otherwise would require lots more asembly instructions.

Re:Miguel is dead!

[ron_ivi]

Parent wrote: "I'm not a C++-hating C bigot "

The more I try to hire good C++ programmers, the more I become a c++hating bigot.

When over 95% of "C++ programmers" that I interview can't even answer this: straightforward question about virtual methods, I see no hope for the language. It seems most people who claim to be C++ programmers just say that because they use a C++ compiler and stick their functions in objects.

The future I expect is a mix of a higher-level language (Java, C#, perhaps even Python) and C. I can think of cases where C is the right tool for the job (small memory embedded systems), and I can think of cases where Java/C# is (large scale enterprise software); but I don't see C++ as the right tool for either job.

Re:Miguel is dead!

[Frymaster]

I can think of cases where C is the right tool for the job

but, really, there aren't that many of those. here's a statement you can argue about:

"usage of c and it's derivatives (c++, objective-c) is more likely create buggy and insecure software. the performance gains from these languages does not make up for this loss of stability"

seriously. direct memory management is a dangerous thing and isn't necessary for 90% of the software written in c-like languages. if you're writing device drivers or an os or whatever, it's important... but does sendmail really need to be written in c?

would it be more secure if it wasn't?

Re:Miguel is dead!

[kdart]

Yep, I read your link. I have a philosophical difference, however. If someone has to know all that esoterica to be a "real" programmer than it's the language that is broken. That language just begs for problems (and bugs).

Re:Miguel is dead!

[Anonymous Coward]

A man is a good programmer or he is not a good programmer. He is not "a good C++ programmer". Whether he has learnt some esoteric nuance of a language does not make him a good or bad programmer. A particular language is just an interface to the act of programming, like a GUI is an interface to the act of using, and if it does something surprising, the problem is likely with the language, not the developer.

Re:Miguel is dead!

[Jordy]

When over 95% of "C++ programmers" that I interview can't even answer this: straightforward question about virtual methods, I see no hope for the language. It seems most people who claim to be C++ programmers just say that because they use a C++ compiler and stick their functions in objects.

I'd like to point out that there is absolutely nothing wrong with writing "C with Classes." It makes a whole lot more sense than trying to fake it with structs and function pointers with only a minimal loss of efficiency. Certainly replacing all char pointers with strings would make up for the inefficiency with the huge amount of time saved debugging. Just because operator overloading, templates, references, etc. exist doesn't mean you have to use them. I mean, no one puts a gun to your head and forces you to use Swing in Java just because it is standard.

Second, that question should have been rather simple to answer for anyone who was aware that the std::complex template has a copy constructor with an implicit conversion. Not everyone knows that simply because complex isn't exactly the most widely used thing in the world. If you are going to ask that quesiton, I'd recommend defining a template in the code to make it obvious.

The language features in C++ can be abused, but a huge number of errors I see are from misusing APIs and none of the high level languages like Java are going to make sure you are paying attention to what you are doing.

Re:Miguel is dead!

[Pete]

ron_ivi:

When over 95% of "C++ programmers" that I interview can't even answer this: straightforward question about virtual methods, I see no hope for the language.

Seriously dude, that is most definitely not a straightforward question. It's carefully designed to test your understanding of some of the subtle tricks and traps of C++. And one of the "tricks" it uses to mislead you has nothing to do with inheritance, rather with implicit casting.

I must admit I got tripped up on it though, mainly because I'd forgotten the distinction between hiding and overriding. The question is nasty, ugly, and was intended to be as confusing and tricky as possible... and any programmer that writes code like that (especially something involving changing the default value of a parameter for an overriden method) ought to be shot.

BTW, if you're asking questions out of GOTW as interview questions, I can only say that that's pretty nasty - except perhaps as a final round question for some of the really cocky smart-arse types :-). Or if you're looking for a seriously hard-core C++ expert, in which case you'd probably expect him/her to have read all of GOTW and Sutter's books and know all the answers off by heart anyway.

It seems most people who claim to be C++ programmers just say that because they use a C++ compiler and stick their functions in objects.

There's certainly a depressing number of such people about - though nowadays they're mainly moving into Java, a language better suited to their limitati^Wcapabilities. And I say "Hooray!" to that :).

Pete (who has for the last two months been maintaining/debugging/adding features to a 1998-era MSVC6 project, written by a guy who really had no idea about C++. Sigh.)

Re:Miguel is dead!

[kahei]

I know a dozen other people have probably already said this, but if you think that question measures whether someone is an effective C++ programmer then the problem is with you.

It's a neat trick and the last bit (where default params are resolved based on static type while method calls are resolved based on dynamic type) fooled me. But it has nothing whatever to do with a software engineer's job, unless that job is writing C++ compilers.

Put away your book of trick questions and hire a C++ programmer who understands patterns, algorithms, project management, requirements, and important standards and protocols.

Re:Miguel is dead!

[RisingSon]

Interesting points.

That link is a good puzzle but I don't know how well its going to help you find good developers in an interview. I remember I was about a year out of college when I discovered such details of C++. I was like, "WTF? I'mn not overloading that one...why am I here?". About 5 minutes and 1 small program resembling the one in your link I had my answer and moved on.

So was it that one year of experience you would be seeking? Or would it be my problem solving skills that when my code doesn't work I'll find the answer quickly and consistently? When I interview, I try and discover one's problem solving abilities, not one's specific knowledge. It obviously all depends on what role you need filled.

A C/Java world with no C++? Most of the software I've written in the past few years has been C++ with lots of C influence. Lots of functions, few objects. I probably have more structs in cvs than classes. Its really nice to have my function/struct style code and my message/class style code coexist freely.

For example, I have a system I wrote that the backend database is all in shared memory. I played with classes in shmem and I put some templates there, too, with the Allocator. In the end, I chose structs because I was able to exploit their fixed size to make things quite speedy.

But now I have a Phd physicist that needs to do some serious matlab-style maxtrix maninpulation of my struct-based data. And he can 'kind-of' code. It was nice wrapping up access with some matrix classes so it all looked more like matlab to him.

I think there is still value in having all that in one language, even though it can make it an ugly one.

Re:Miguel is dead!

[Gr8Apes]

So that's where all those idiot^H^H^H^H^K inexperienced OO java programmers come from! I was wondering.

FYI, a bad OO programmer in C++ is most likely going to be a bad OO programmer in Java as well, except the language will most likely prevent them from taking down the entire machine.

Good OO programming is good OO programming, regardless of the language.

Dealing with MS patents ....

[konmaskisin]

1) Rave about C#
2) Convert everything to run on Mono C# .Net
3) Get sued bad by the world's most deep pocketed software publisher except that unlike SCO vs. Linux this time the evidence is on the side of plaintif.

Re:Dealing with MS patents ....

[Valar]

How?

MS has given everyone license to implement every part of C# and .NET except for the Microsoft namespace of the class library. They have given up their right to sue people who use .NET or make .NET compatible implementations. The evidence isn't on their side.

Hint: ECMA.

Re:Dealing with MS patents ....

[foandd]

They have given up their right to sue people who use .NET or make .NET compatible implementations

Christ man, quit sniffing glue before you hurt yourself. RAND doesn't mean they give up the IP; it means they agree to license it to all parties who wish to license it, and that all parties will pay the same rate.

I used to work for a company who got their heavily patented IP accepted as a standard. It has since become ubiquitous it its field. It was accepted as a standard under the same type of RAND terms which are being discussed here.

The numbers have likely changed since then, but when I worked there if you wished to implement it, you had to pay a $25,000 up front fee and a $70 fee per shipping unit. If you implemented without paying those fees?

Hint: their lawyers would be up your ass with a flamethrower before you had time to blink.

What a choice...

[eidechse]

I'm not sure I dig the idea that the future of app development comes down to a choice between MS and Sun.

Parrot/Perl6

[Freedom Bug]

IMO, I think Parrot will become the .NET equivalent in the open source world: a language independent VM and a huge standard library. I can't wait to have access to CPAN from my Python programs.

Sure, C# is a lot nicer than C, but Python & Ruby are a lot nicer than C#. If you're going to give up the predictability of C/C++ for a VM, garbage collection, et cetera, why not go all the way up to dynamic execution?

Bryan

Re:Parrot/Perl6

[bcrowell]

Parrot has at least two big things going for it:

1. It's fast. It was designed from the beginning to be fast. They're not leaving the optimization for later -- the early releases are already super fast. IIRC, some languages compiled into Parrot are already faster than their previous implementations.
2. It's immune to the kind of MS mischief that .NET will undoubtedly have to struggle with.

Netcraft confirms: C is DYING

[Joey Patterson]

It is official; Netcraft confirms: C is dying

One more crippling bombshell hit the already beleaguered C community when IDC confirmed that C market share has dropped yet again, now down to less than a fraction of 1 percent of all servers. Coming on the heels of a recent Netcraft survey which plainly states that C has lost more market share, this news serves to reinforce what we've known all along. C is collapsing in complete disarray, as fittingly exemplified by failing dead last in the recent Programming Language Usage Audit.

You don't need to be Scott McNealy to predict C's future. The hand writing is on the wall: C faces a bleak future. In fact there won't be any future at all for C because C is dying. Things are looking very bad for C. As many of us are already aware, C continues to lose market share. Red ink flows like a river of blood.

C++ is the most endangered of them all, having lost 93% of its core developers. The sudden and unpleasant departures of long time C developers Mark Markup and Sally Smith only serve to underscore the point more clearly. There can no longer be any doubt: C is dying.

Let's keep to the facts and look at the numbers.

C++ leader Theo states that there are 7000 users of C++. How many users of C# are there? Let's see. The number of C++ versus C# posts on Usenet is roughly in ratio of 5 to 1. Therefore there are about 7000/5 = 1400 C# users. C++ posts on Usenet are about half of the volume of C# posts. Therefore there are about 700 users of C++. A recent article put C# at about 80 percent of the C market. Therefore there are (7000+1400+700)*4 = 36400 C++ users. This is consistent with the number of C++ Usenet posts.

Due to the troubles of Redmond, abysmal sales and so on, C++ went out of business and was taken over by Microsoft who sell another troubled programming language. Now C# is also dead, its corpse turned over to yet another charnel house.

All major surveys show that C has steadily declined in market share. C is very sick and its long term survival prospects are very dim. If C is to survive at all it will be among language dilettante dabblers. C continues to decay. Nothing short of a miracle could save it at this point in time. For all practical purposes, C is dead.

Fact: C is dying

Exactly

[SuperKendall]

Exactly, there will be "kind of" a functional language - which has to use the same libraries everyone else does, which are all just like the Java libraries. So people using this pseudo-functional language will be hard-pressed to really see the advantages of a functional language as you would if you had a real function language with a set of libraries as broad as that offered by Java.

I formed this opinion long ago, just around when C# first came out, when I read an article by one of the founder of Eiffel talking about how Eiffel# would work (I wish I had not lost the link - I can no longer find this article). He saw opportunity at that point to gain new Eiffel programmers. But I saw only a tar baby, where the longer you worked with the system the more you just said "Well, all I'm doing is calling C# libraries with this weird syntax so I might as well make life easier by just using C#". .Net and C# is the perfect vehicle to migrate users of all other languages to C#, not really to make life easier on people who want to work in one of the non-C# languages.

Re:Exactly

[chgros]

Exactly, there will be "kind of" a functional language - which has to use the same libraries everyone else does, which are all just like the Java libraries. So people using this pseudo-functional language will be hard-pressed to really see the advantages of a functional language as you would if you had a real function language with a set of libraries as broad as that offered by Java.
I heard F# was supposed to be based on OCaml.
OCaml already comes with a nice set of libraries, not to mention parser tools (which I heard were missing from C#). It even has bindings for SDL!
And I don't know why you call it "pseudo" functional.
Although I agree on the fact that the syntax for .NET bindings would probably have to be weird.

Miguel's fantasies

[A nonymous Coward]

The more he slides under Microsoft's spell, the more I distrust his pronouncements. It's as if he thinks he alone has discovered the magic potion that will allow him to befriend Microsoft with groveling and flattery such that they will actually respect him and not pull the rug out from under him. Why he should be the one exception in all their history is beyond me.

I have pretty much come to the conclusion that if Miguel says it is so, it ain't.

Hell yea PyGTK!

[3cents]

I've been using pygtk on and off for the last year or so and i've got to say that when used in combination with glade it's a real dream to work in. I think if a Visual Basic like IDE could be created for those who are afraid of the deep end of text editor, command line tool development we could have a real open source Visual Basic killer on our hands. And i'll take that over C#/Mono any day of the week.

slashrank

Give me a break

[Anonymous Coward]

C is dead? What is he smoking. It's going to be around for a long long time. Will enterprise apps be developed with C, sure it will. Will C# get used for "enterprise apps". Well smallish non-transactional, horizontally scalable apps sure. .NET is still lacking when it comes to a proven, robust messaging server and transaction monitor. Don't say COM+, unless you've actually taken time to read the full disclosures hosted on http://www.tpc.org/. Stock COM+ transaction API is ok, if you want to handle update/commits in an async fashion. But don't expect it to be scalable in a massive or medium sized trading system. It will get better with Whidbey, Indigo and Longhorn, but there's still a ton of problems with Windows at the kernel level that keep it from scaling well for hardcore backend applications.

don't bother mentioning .NET community sites that host on .NET or Match.com. Those aren't transaction heavy or trading systems. Glorified webpages that serve up database tables are simple. A decent developer can build it. Building a transaction application that can handle 500-1000 moderately complex transactions (update/insert) per second is hard. .NET can't handle it, or atleast I all the cases I've heard of in the financial industry failed miserably. If you're in this industry you already know. If you're not, you're probably saying "what bs, .net can scale just as well."

Yeah, try to build one in .NET and tell me how many servers you're going to need. then tell me how you would run real-time analytics on all the rows in a given table when you've partitioned the table across 8 machines and each has 8 million rows. When the CTO says, the query has to take less than 30 seconds, what will your answer be?

Alternative mono names

[bangular]

After a long time searching I found these alternative names they considered.
Foot gout
Genital Herpes
Oily discharge

That is where you are mistaken

[SuperKendall]

.Net is the platform in which C# is the closest natural representation of what the bytecode is like. As such, .Net is a vehicle for C# and a distorted reflection of all other languages.

Yes, .Net is the VM and C# the language. But as Java has shown, it's not really so easy to keep the two wholly separate. They are interdependent to some extent.

Re:That is where you are mistaken

[Rob Riggs]

.Net is a vehicle for C# and a distorted reflection of all other languages.

A very good illustration of this is in the attempt to create a Python.NET implementation by the folks at ActiveState. The report on the lessons learned is rather enlightening.

Re:That is where you are mistaken

[de Selby]

"There is no support for some Python features that some programs will require. Most of these are
fairly obscure, but are a limitation. Examples include:
- String formatting
- Core language features, such as long integers, complex numbers, built-in object methods
and so forth.
- The standard Python library."

Are they joking?

Re:That is where you are mistaken

[srn_test]

I think so. They go on to mention that other missing features are things like multiple inheritance, and the ability to reference self in the constructor, and to avoid calling super-classes constructors.

In the end, though, they come out with something about how "language neutral" .NET is.... They seem to be Windows pukes, though - they refer to .DLLs all the time, rather than the more general "shared libraries".

There's certainly no way pretty much any Python code I've written (and I've written a lot) would run on .NET, given these limitations.

Important comparisons between .Net and Linux

[Bodhammer]

From Infectious Mononucleosis

1) When people think of ..., or "mono," they often think of extreme tiredness as one of the major symptoms associated with this illness.

Ditto

2) Loss of appetite and generalized weakness also may be present, especially in adolescents. Nausea, hepatitis, jaundice, severe headache, stiffness, chest pain, and difficulty breathing can occur in some cases. A pink rash can occur all over the body in children who have been treated with ampicillin or amoxicillin.

ditto again!

3) Some may experience extreme fatigue, staying in bed for more than a week because they feel too weak even to walk around the house.

We're three for three (esp. after a failed Windows update session...)

4) 'Mono' is generally a self-limiting disease, which means it goes away on its own in most cases. Occasionally mono can cause complications

Need I say more...

5) Epidemic outbreaks in hospitals and workplaces have occurred.

` My biggest fear! Remember kids - practice safe computing!

Why Mono Will Fail

[cryptoluddite]

There is almost no difference between C#/Mono/.NET and Java, but almost no Linux developers write in Java. Check out your distribution's packages and you'll almost see more JVMs than Java apps. And for some reason Linux developers avoid Java like the plague, even though it's got a godzillion features that make everything so much easier (garbage collection, huge consistent class library, security, etc). Put in a GTK or QT library interface instead of the slow and huge Swing (that Smalltalkers foisted on Java) and you're golden -- there's every reason to use Java, especially for applications.

The Linux culture has so far prevented Linux from taking the next step. Just look the (essentially) complete lack of interest in gcj (gcc open-source java). Just look at the slow pace of Mono. It isn't goind to happen anytime soon, unless the Linux app community wakes up and sees the future. Yeah, 10 years from now we'll still be doing manual memory management. Sure...

Re:Why Mono Will Fail

[primus_sucks]

Put in a GTK or QT library interface instead of the slow and huge Swing (that Smalltalkers foisted on Java) and you're golden -- there's every reason to use Java, especially for applications.

SWT is a crossplatform UI toolkit that feels like a native app, unlike Swing. In Linux it's just a wrapper over GTK or Motif.

Re:Why Mono Will Fail

[LarsWestergren]

There is almost no difference between C#/Mono/.NET and Java, but almost no Linux developers write in Java.

Not quite true. Current Sourceforge statistics by language:
1. C++ (12967 projects)
2. C (12955 projects)
3. Java (11446 projects)
4. PHP (8607 projects)
5. Perl (5388 projects)

There is a vocal group of Java haters on Slashdot.

There is a bigger group who is a bit wary because even though the Java Community Project exist and IBM and others are allowed to develop their own implementations, the licence is not GPL and Sun could screw things up if they turn evil or desperate.

Check out your distribution's packages and you'll almost see more JVMs than Java apps.

Yes, because lincence restrictions prevent them from including Suns java. The open source packages are not mature enough to run all apps, so it's pointless to include those apps. The open source implementations of the VM are impressive, but the java standard is developing too quickly for them to keep up porting all libraries. Just like .Net will evolve too quickly for Mono to keep up. Only in that case it is guaranteed MS will deliberatly include things that will break Mono.

there's every reason to use Java, especially for applications.

Yes, and I believe a huge reason for the upswing of Linux in businesses is the great combination of Linux and Java.

Do NOT help Microsoft kill off Java with their .Net clone, when they are done with that they can concentrate on the single remaining target.

This is the time...

[Dr. Sp0ng]

... to attack Microsoft on the .NET front. We have all the components of a powerful, truly cross-platform development platform; they just need to be packaged in a way which can compete with the cohesive API that is .NET.

Now, I'm not trying to start a flamewar here, so I'm not going to insist on specifics here, but what needs to be done is to create essentially a "distribution" of a powerful language (I'd use Python), a good cross-platform widget set (I'd use Qt, were it not for the license issues - personally I don't blame Trolltech for the licensing scheme, as $2500 is chump change to any real development house anyway, but in this case it would hinder adoption. Maybe wxWidgets then, or GTK+ configured in a way to appear like native widgets, but gah, GTK+ is such a hideous API), a selection of useful Python (or whatever) modules, and so on.

Like I said, this is nothing that didn't exist before, but it should be packaged in such a way as to provide a standardized environment on every platform that applications can rely on. Come to think of it, rather than use a specific language like Python, maybe the whole thing should just be wrapped around Parrot.

Anyway, the specifics don't matter. It's the idea of having a good competitor to .NET with a standardized set of support stuff on each platform, with a single installer to support the entire shebang. Why can't this be done? If I had the time, I would, but I don't. But if there was one place the open source community can take on Microsoft, it's development tools. And the time to do it is before .NET gets too entrenched.

Re:C is dead?

[maraist]

C is inferior, yes. It's hardly dead though.

I think C deserves a little more respect than a blanket comment like this.

We still have assembly languages today for several reasons, even though it's been "dead" since the 80's. Driver writers, compiler writers, and high performance inner-loops of scientific apps NEED snippets of assembly.

Thanks to the design goal of c (as the paint on the metal), you can now program in mixed c/assembly and thus never have to actually write anything in raw assembly files anymore. But you still write with the intention of generating specific sections of assembly.

The people that write c are those that are resource conscious. I am not aware of ANY other language that is allows resource management as well as C. "performance" is just one resource, and I know many argue that JIT's counter-balance the performance advantages of optimized c. I still refuse to belive JIT's have ever approached -O3 performance, but am willing to concede that a well written VM app can perform acceptibly.

ButMemory is another key resource management position that sometimes requires hard computer science to derive workable solutions.. When you abstract how resources are utilized (via a VM), then it negates the value of such hard computer science. Note that this ONLY applies to a particular problem space. Though, a non-trivial problem space to be sure.

The point is thus that the classic paradigm of "the right tool for the job" is essential here. There are MANY problems that are best done with low level, very concise languages like c. Most OS components, VMs, or fast-duty-cycle applications should very well have core work done in c/assembly.

The obvious other end of the stick is managing large amounts of code, for which assembly need not apply, and c is indeed becoming a distant memory as I think you are implying. Yes, web-services and many such fast-to-deploy applications transcend c, but don't be too quick to write-off the value in having a generation of computer scientists that are not well enough versed at writing low level applications.

One thing you need to know about DeIcaza..

[Bowie J. Poag]

Hmm.

The initiative it takes to back-engineer the .Net frame work, while admirable, is a wee bit misguided, i'd say. You have to understand what's really at play here.

The whole idea of chasing after a proprietary standard like .Net (and treating the development process like some open source cat-and-mouse game) is pointless. It accomplishes nothing but an at-best "acceptable" knock-off.

If Microsoft doesn't want the .Net framework available on other platforms, they're not going to make it available. Then on the other hand, if they find it's to their benefit to do so, they will make it available. It's as simple as that. Suppose you actually make it to the point where you have an equivalent framework to offer. Surprise, Microsoft releases .Net for Unix, and the entire effort is null and void. Your pride and joy is now a footnote, and a deprecated one at that. Realistically, how many people are going to feel sorry for someone who's stance reads, "Damn those fuckers in Redmond for making their framework available in Unix!" ?

Unfortunately, Miguel seems to have a fairly long track record with this sort of Microsoft-chasing. You can run along and play catch-up all you like with other technologies, but, it'll be just like it is with Gnome ---- With Gnome, you're guaranteed nothing better than a perpetual second place finish. Gnome amounts to a Windows wanna-be, instead of a Windows-killer, when it didn't have to be that way. Miguel made it that way.

Look...If you HAVE the talent to do something better, for God's sake, do it. Don't waste your time (and other peoples time) churning out flea market knock-offs of worthwhile products. I'm certainly no fan of Microsoft..I just call it like I see it. .Net needs not the taint of open-source muckery.

The peril that is .Net

[SuperKendall]

You might want to look at Nemerle (www.nemerle.org)

You have brought forth the perfect example of the danger of .Net for all other languages. Take a look at one of the first paragraphs describing Nemerle from a PDF I found on the site:


Of course, the framework is strongly object-oriented and primarily focused on traditional object-oriented and imperative languages. Therefore ports of the existing functional languages to the .NET did not fit in as well as, for example, C# [5] does. Addressing this issue was the main idea behind the design of Nemerle. In comparison to Haskell [3] or SML [2], Nemerle is not a pure language in the functional sense, allowing the programmer to create completely object-oriented and imperative programs. This makes Nemerle a good transition language for people with C-like imperative and object-oriented languages background. They can take advantage of the language imperative features until they gradually learn how to program in a functional fashion.

No, not quite a pure language and you have to use the C# libraries. Again, a pale reflection of a real functional language with libraries that do not encourage functional use.

But the real problem is the last sentance - "until they gradually learn to program in a functional fashion".

You and I know how programmers are. We are built for laziness. I am that way myself. So what would happen if we put twenty programmers in a room, ten of which came knowing C# and ten knowing a functional langauge, and had them program in Numerle for a year? I'll bet 18 of those twenty would just be using C# by the end (even knowing how obsitinate supporters of functional languages such people are!), and the C# people would not have picked up a lot of functional knowledge along the way. Because knowing how C# works gets the bulk of work done in the system (as like Java, the bulk of work is done by a rich set of libraries), you will evolve to working in a C# style even in the confines of a functional language!

How many Eiffel# programmers are there out there already? How many people have used .Net as a "Stepping stone" to learn the other languages offered by the .Net platform?

Generics

[SuperKendall]

Uhm, why would MS do this? .NET is designed to be and sold as a "cross platform" solution--a real way to do Java's "write once, run anywhere" line.

Introducing a crippling change for Longhorn would not only hurt this goal, but it'd bring down the rath of a few regulatory agences all over again--and piss off everyone using .NET and not Longhorn.

Java's write once, run anywhere is pretty real - it's harder for GUI's of course and works best for server side stuff (who even thinks anything of compiling and testing in windows or linux and just moving Java compiled code up to Solaris or some other system?).

But to address your point above, the introduction of Generics will change the .Net platform significantly and require new .Net installs - code built using Generics on the future .Net platform will not run on older versions of .Net. Now they get some nice features for that tradeoff, but it is just one of the changes .Net is introducing that will require updates in Mono. And Microsoft has indicated a lot of other API's will change in the future with Whidbly(sp?) as far as libraries go.

C is dead

[corian]

Great! Now there's only 25 letters I have to teach my kids.

Re:COMPLETELY different

[mhesseltine]

Operating systems are NOT, I repeat, NOT in the same boat as operating systems. OS/2 was a "better" windows, but windows already had taken hold of the market. Mono is just allowing current linux developers to take advantage of a very powerful and capable language specification which, BTW, will also allow many people to stop needing to use windows.

First, I would say that Operating systems are in the same boat as operating systems. Maybe this was supposed to say "Operating systems aren't in the same boat as programming languages"?

People aren't trying to build a better .NET; we're trying to build a .NET. Is there a .NET for linux?

Are you implying we shouldn't use .NET on linux merely because .NET exists in windows? I think that's not a very good argument...

I think the implication is that, if you trust Microsoft to set the standards, you'll spend all your time and effort playing catch-up to them when they change to a new standard.

For an example of this, look at the Samba project. Yes it's been successful, but look at how difficult some of the advanced functionality was to get correct (Active Directory for starters)

Mono will always be a second-class citizen in the .Net world, precisely because it isn't supported or backed by any truly large player that can hold some influence with Microsoft.

Enthusiasm

[Brandybuck]

So far, there's a surprising enthusiasm for Mono and C#

Latebreaking news flash! A reporter attends a Mono lovefest and discovers that the attendees are having a lovefest!

New Slogan: Catch Mono!

[Captain Irreverence]

Mono is a DotNet wannabe.
DotNet, of course, is a Java wannabe.

Why not just use Java? Oh, that's right:

Microsoft - Duplicates anything they don't own and control

Open Source Movement - Duplicates anything someone else owns and controls

C not dead, .NET/C# not interesting

[Anonymous Coward]

If I need to write a daemon that handles 250,000 hits/hr, I write it in C. Hands down, C is the only language I know where I can control memory and CPU usage. The only language where I can predict what will happen with some accuracy. If I write a program in Python or Java or whatever, it starts consuming RAM, it starts context switching at the wrong times, etc. On a recent Python project I had to actually go into the Python memory allocator (written in C) and do some hacking, because it didn't work right for my project. So obviously C is well and alive.

If I need to write a quick script, but make sure it's clear enough to understand next week, I use Ruby. A lovely, well-designed language. Why bother with compiling and linking? Go dynamic.

If I want to use an awesome development environment, I use Mac OS X. Aside: sometimes I wonder if people who rave about the latest scripting language or the latest Windows dev tool, have ever actually *used* NeXTSTEP/Cocoa! It's still light-years ahead of what most people use. Look how many examples in the Gang of Four patterns book come from NeXTSTEP!

And, if I need to write a .NET program on Windows, I ... USE WINDOWS! Maybe I'll "play" with Mono, but at the end of the day the only reason I'd use Mono is for Windows interop. Not to replace Windows.

C# is a nice improvement on Java. Which isn't hard to do.

Like a lot of folks I look at Mono and I just don't get the point of it. It's Microsoft's technology, it's always going to be playing catch-up... to a mediocre environment.

Give me .NET's dev tools and IDE for Ruby, then I'll get excited.

Mono?

[RoadkillBunny]

What does Mono stand for? I heard it stands for 'Microsoft, Oh No'

Danger!

[dmiller]

I respect Miguel, but I think he seriously underestimates the risk posed by MSFT's patents in this area. Quoth the article:

Microsoft has granted a license to use this technology under so-called "reasonable and non-discriminatory" terms.

"Reasonable and non-discriminatory" (RAND) does not imply "free". RAND was the proposed licensing requirement for W3C patents that was howled down by the community.

Given that MSFT is willing to finance SCO to use arguably illegal tactics to destabilise and discredit free software, who would expect that they are above enforcing a small fee for every patent needed to implement Mono? They needn't do this immediately, in fact it is in their interest to wait until the technology is widely adopted, so they can slug everyone at the same time. Note that the usual legal defences against "submarine patents" won't work either if the terms have been disclosed to be RAND all along.

Re:Danger!

[hayden]

"Reasonable and non-discriminatory" (RAND) does not imply "free".

RAND licencing usually means reasonable and non-discriminatory" for business. What MS can (and probably will) do if Mono gets any traction is say, "We'll licence the right to use all these patents for 10 cents per install". Perfectly reasonable. Perfectly non-discriminatory. Kills any open source implementation dead.

Game over.

Why should we bother?

[bonch]

Well, if we can get to the point where we can do things like those shown in this Longhorn video, development for Linux will definitely increase.

By the way, the video is worth watching not only for the cool technology but also to hear two Microsoft programmers ragging on each other. Totally opposite the kind of atmosphere you'd expect at Microsoft, I suppose.

Re:Why should we bother?

[JabberWokky]

The .exe extracts a large .wmv file. Perhaps someone could mirror?

Just use cabextract. It's bundled with most distros. Pretty common linux tool.

--
Evan

Gmone needs to distance itself from Mono

[oob]

We have Gnome because Trolltech refused to release QT under a free license in the early years.

The prevalence of KDE at that time showed us that a large proportion of the F/OSS community would cave to non-free software if it was convenient.

Until recently a mutually beneficial relationship existed between Gnome in the community and Ximian as a commercial provider, analogous to the relationship that exists between OpenOffice.org and Sun's Star Office.

Now that Novell owns Ximian (OK Novell are currently our friend, but that may not always be the case) and Miguel is exposing the community to litigation from Micro-Soft down the track through the introduction of non-free technologies (and procedures, methods, etc. which are likely to be covered by upcoming patent laws) the GNOME project needs to distance itself from the disruption that the Ximian guys represent in order to avoid contamination.

Miguel has a solid reputation and has done a lot for the community, but in this case he is either (best case) a fool for thinking that Micro-Soft can be trusted or (worst case) a quisling. Either way, it's barge pole time.

What a load of rubbish

[Chuck Chunder]

Surprise, Microsoft releases .Net for Unix, and the entire effort is null and void. Your pride and joy is now a footnote, and a deprecated one at that.

Sun produces Java for a whole bunch of platforms. So does IBM.
Intel have a C compilers on a bunch of platforms. So does GNU (and Sun, and Microsoft).

Software doesn't suddenly become useless simply because another vendor releases a competing implentation on the same platform. If MS do release a .NET implementation for Linux then mono will still be valuable. Licencing terms will obviously be one area they compete on. No doubt there will be others.

Re:No language can replace C ...

[Alex_Ionescu]

The first programming language(s) were coded in ASM, Assembly Language. Although ASM is a "language", it doesn't need to be translated and compiled like something like C. Each operation corresponds to opcodes, which are usually one or two bytes, that the CPU can directly read and interpret. Once powerful and stable C compilers (written usually in ASM) started to appear, new languages were usually directly written in C, but always containing some ASM.

The main problem that confuses people (such as the argument that C# was written in C#) are the IDEs (the Integrated Developpement Environment), which is merely a graphical frontend to do your work. So yes, the VS NET 2003 UI was made in C#, but the compiler for C# was written in C and ASM.

I hope this makes sense...it's 2:30AM...lol

Best regards,
Alex Ionescu
Relsoft Technologies

Quoted way out of context!

[1in10]

Miguel said, "C is dead", sure.

But how about some context. This is what he actually says: "To me C is dead. Except for the JIT!" (emphasis mine).

He's not saying C is gone, that nobody is going to use it anymore. He's saying that he's not using it anymore, except to write the JIT compiler for mono.

So the quote is accurate, but the meaning is completely changed once you put it into context!

Once you use Eclipse you'll understand

[SuperKendall]

Eclipse is way more useful hat Vs.net (which I have used).

For one thing, imagine having a compile error and just saying - "Please fix". Often Eclipse gets this right in one, or can offer a range of choices (not all of them right all the time, but even so...).

And then there's great refactoring support that is really lacking in Vs.net.

With Java you actually have a range of editors that all offer more advanced features than Visual Studio. In IDE's you have Eclipse as I've said, or Idea, or JBuilder (expensive but very very easy). To compete with .Net for server side ASP development you have Weblogic Workshop which can help define things like flows between pages graphically.

And those are all if you have to have IDE's - developing Java without an IDE is I think simpler than .Net. It's probably a bit of a wash as both sides have Ant, but Java has a wide range of ant tasks defined already.

In every regard, .Net is a platform furiously trying to re-write every useful Java tool - and almost succeeding. Can it keep up the pace though?

Out of Context

[superyooser]

The submitter quoted Miguel out of context. Miguel said:

" To me C is dead." (emphasis mine)

I take that to mean, "As far as the world of Mono/.Net goes, C is dead."

Re:C "Dying", Scripting Languages, Cross Platform

[Jellybob]

Mono is also write once, run anywhere - the .NET platform is designed to be cross platform (even if Microsoft were hesitant to make use of that).

I have written programs in C# on my desktop, compiled them with mcs (the c# compiler with mono), and then e-mailed it to a friend running Microsoft's .NET runtime on Windows, and it ran without any problems (obviously you have to make sure you don't use the Mono APIs if you're doing this).

To push it to further extremes, that person could then upload the same compiled file to their PDA, and *still* be able to run it, so long as they have the runtime installed.

How much more cross platform do you want?

and of course

[pyrrho]

half that esoterica wasn't even implimented while C++ became popular.

Because functions declared in structures is 90% of what C programmers wanted out of it at the time. Having constructors and destructors was also quite nice. And that is enough to use C++ right there.

You are not supposed to use all the esoterica.

hey look

[pyrrho]

I can write a web server in bash in one line! AND it installs itself.

apt-get install apache

verily... bash has gazillionized my productivity... it used to take me over six months to write a decent web server.