Slashdot Mirror
Mono and dotGnu: What's the Point?
joeykiller writes "The Register features an opinion by Neil Davidson, asking 'Mono and dotGnu: What's the point?' Some of the points he raises may seem irrelevant for open source supporters (like why make a C# compiler while Microsoft's is free anyway), but others are thought provoking and maybe a little bit controversial. You may not agree with his opinions, but it's an interesting read anyway."
It reads like a troll. A software company directory who doesn't "get" why you need to
bootstrap a compiler.
Although their time might be better spent in designing a true alternative to Java and C# instead of a copy that allows you to write a GNU application that runs everywhere, it's hard to fault Mono for recognizing a market niche and running with it. For example, maybe they'll make C# work on Linux embedded devices where Microsoft wouldn't go?
Someone should tell him that there are other processor architectures than just x86, processors that aren't supported by MS ...
codegolf.com - smaller *is* better.
... doesnt always mean it makes sense from a resource point of view or from "the big picture". But that is the price with people giving up their own time.
I, myself, am happy to have the chance to sample some of this work for free. Who am I to judge since I'm not paying?
We know that Slashdot posts these kind of articles as bait for the meatheads who think that being against Micro$oft makes everything they say correct. Microsoft bashing is no substitute for thinking. Be free! Think, think!
As a .NET developer, frankly, I don't care what the motivations of Mono develoers or dotGnu developers are. Maybe I should be, but I'm not. I'm building an open source project in .NET and I want Linux, BSD and Mac OS X support (the latter two, hopefully with the help of SSCLI), and frankly, whatever other platforms I can include.
.NET is a really nice development environment. As much as I don't care for Microsoft, I have to admit that since I adopted C# about a year and a half ago, my production has roughly tripled, maybe more. I've never had ANY technology have that kind of impact on my development before, unless it was the reverse (making me 3 times LESS productive).
I don't want to use Java. First of all, I've never used it to develop software. Second of all, every user interface I've ever seen done with Java stinks. Maybe I've been seeing bad examples, but the windows, buttons, and other contols of the Java apps I've seen have an old fashion look and feel to me and I don't care for it. My personal opinion, but for me, that counts for something.
So, whatever the motivation of Mono or dotGnu, I simply want to develop my cross-platform C# apps. That's MY motivation, and that's what matters to me.
We read with interest the piece from Neil on the purpose of Mono, and
.NET software; You would think that the use of Mono would help
I wanted to clarify a few things, because Mr Neil does not seem to
have looked at the Mono Roadmap, nor tried a recent release, since
code signing (authenticode and strongnames is implemented, remoting is
completed (soap, binary, http, tcp transports) and most of the
side-by-side assemblies work is done, and will be part of 1.0).
The Mono Roadmap (www.go-mono.com/mono-roadmap.html) contains the
release time frames for the various features of Mono and will help him
and other readers understand what exact plans are: no speculation, and
no half-cooked facts.
I am surprised by the motivation to do so little research on our
project given that Mr Neil is the technical director of a company that
sells
him reach customers using Linux, using mainframes or MacOS X.
Mono is based on the ECMA 334 and 355 standards. We like the C#
language and its runtime (as does Mr Neil's company) because it
increases our developer productivity, reduces the time to market of
our new products, this despite the fact that we do not implement Code
Access Security, which will only be used in embedded situations, a
segment that we are not ready to address in Mono 1.0.
We want to improve the productivity of developers in Linux, mainframe
and OS X developers by brining this unique platform to other
platforms. Just like Borland, SGI, Sun and IBM provide compilers,
runtimes and tools for other languages, we provide such a piece for
C#/.NET.
Mr Neil does not seem to understand why bootstrapping a C# compiler is
important, so let me explain this in terms he would understand: it is
important because:
* Using C# to write a C# compiler means that it improves our
development speed.
* To be able to harvest the benefits of productivity of C# on
Unix, we need a bootstrapping system.
* It allows us to write software on Unix without and be
self-sufficient to develop software as opposed to require
a Windows machine to develop software, and another to run it.
* It means that we trust our technology enough and it is solid
to the point that a relatively complex piece of software not
only runs, but is binary-compatible with the Microsoft
runtime.
Mono's objectives are not "To break Microsoft's monopoly". We do not
define ourselves in this way, there are more important social causes
to fight. We look at the ECMA 334/335 standards as a solid foundation
to improve Linux and bring more software, more quickly to it, and make
the development process more fun.
There is a lot more about this on:
http://www.go-mono.com/rationale.html
And a few other interviews
WINE is a bad idea in the long run. Sure, it may bring people to Linux but they are still using Windows program on Linux. They never break free of the Windows lock box even though there are (in a lot of cases) software that can do what they need natively.
I agree with quite a lot of this article myself. I've never liked Mono or thought it a very good idea. .NET is an OK platform but it exists solely to keep customers locked in to Windows. A quick look at the Longhaul previews show that it will be next to impossible to write Windows software in the future without embracing .NET. Java remains a real and viable threat to Microsoft but Mono is not since you can be sure that, should it ever become any good, Microsoft will start using spurious patents and other nastiest to kill it off. IF they weren't sure they could do this they wouldn't allow it - what otherwise was the point in spending all that money on a MS Java clone in the first place - they're not going to make any money off it directly.
Despite the fact that this guy doesn't seem to understand or appreciate Mono and dotGnu acheivements, I still pretty much agree with his conclusions.
Let's look at Java. How many professional/hobby/academic Java developers use Sun's SDK?
How many use Gnu classpath with some other VM?
Have you ever downloaded an app or library that was developed and tested under SableVM/Gnu-classpath but not Sun's SDK?
-... ---
While providing freedom of choice might be reason enough to justify a project, practical programmers could be asking: What's the point?
.NET framework on Linux be good enough so that people will actually want to use it? That's a very real question, but it's not the main one he's asking.
Linux, maybe? Mac OS X? Free BSD? I see getting C# programs to run on other platforms as a practical purpose. Later on in the article he acknowledges that you'd be able to run these programs on Linux, but that's more like a throw away concession he makes. He plays dumb in the beginning, and makes himself look silly.
How is making C# a standard on Windows and Linux going to hurt Microsoft?
I think that the people behind the project have better goals than that - namely, getting a particular tool to work on Linux. People use Linux for a variety of things. It would be nice if C# - just another tool - worked under it. What's the big deal?
There is an obvious practical purpose to getting C# programs to run on Linux. The real question, however, is will the
Hey all,
.NET the better. I, for one, am glad to see that the effort is being made and that .NET is not going to become yet ANOTHER MS only technology. If you think Java has merit (and it does), then you can't reasonably believe that .NET has nothing to offer -- they're conceptually the same thing, skewed in slightly different directions. Java is bent more towards security, while .NET is bent more towards flexibility. It makes sense to bring it to Linux: it's useful (really!).
It seems to me a little hypocritical to complain about MONO and dotGNU when there's also WINE out there. What's the point of getting Win32 Apps to run natively under Linux? We'll never keep up with MS adding things to the API...
The point is, the more implementations there are of the CLR for
On the flip side, why isn't anybody complaining that there's an abundance of Java VM implementations out there?
C
The Sun is proof that we can't even do fire properly.
Mono is currently focused on properly wrapping Gtk and GNOME functionality (pure wrappers, without going through Forms). There are already a few apps that use this, and at least one is IMHO a potential killer app (namely Dashboard).
.NET knockoff.
It goes without saying that both Gtk and Gtk# (Gtk wrapper for Mono) work on Windows, too. So you don't lose cross-platform angle, but this does show that Mono is *not* just a
--
I refuse to use
It's a more established framework and a lot more cross-platform.
I think the real point that he's missing is that every project undertaken on Open Source that's a direct response to something that Microsoft is doing is a step in the direction of eliminating barriers to entry. Anything that can be done on an Open Source platform that could previously only have been done in a Closed Source environment is a good thing.
Why make another C# compiler when Microsoft already has one? That's a pretty silly argument. I wonder what Neil thinks of the redundancy among C++ compilers: Borland, Watcom, Intel, CenterLine, IBM, Sun, Microsoft, GCC/G++... More choice is a good thing, people can pick the one that best suits their needs.
Also, whether Microsoft's C# compiler really adheres to an open standard or is eeeeeeeevil closed proprietary software is purely an academic question unless someone makes a non-Microsoft version. Then any parts of the original implementation that are mysterious black boxes, encumbered by patents, or otherwise not really so open become apparent.
Wuba wuba-
The C# languages specification is only part of the ECMA AND ISO standards that have been ratified... As well something called the BCL or Base Class Library was submitted as well. Over 900 classes, namespaces and interfaces.
Sure, Microsoft specific namespaces aren't there like WinForms, WebForms, Data & EnterpriseServices - but you would be surprised as what IS:
- cryptography
- networking (not just sockets, but higher level protocols as well)
- XML
I don't know why everyone in the open source community feels compelled in chasing behind Microsoft technologies, whether it be Mono or Wine. When I talk to people about the benefits of Linux and open source; I always seem to always mention Apache, Perl, and MySQL. I mention these products not because there based on or copied from Microsoft technology. It's because they are innovative open source projects. These open source projects do well not because there open source; but because there BETTER then there closed source counterparts.
.NET our own technologies, and make them BETTER then their closed-source counterparts. That's the only way we win. You cannot win a race by chasing your enemy. You must pull ahead.
We shouldn't lag behind and chase Microsoft's coattails. We should instead innovate; create our own
Neil is the type of person that would argue with the mountain climber about whether there's a reason for climbing the mountain. I imagine him saying, "But Why!?!" For all the other programmers out there, continue honing your skills and techniques even if the apps you create aren't appreciated or widely acknowledged
What's the point of Microsoft making a web browser? They obviously can't keep up with all these standards like CSS, XSL and XUL... I think they should just quit.
It cuts both ways.
Due to the SUN license, no distro would bundle the JDK in the Free version. Kaffe was /usr/bin/java for all of the GNU/Linux distros.
They have no control over C# standards in exactly the same way they have no control over HTML standards: Developers code to the Microsoft implementation, regardless of what the "standard" may say.
To my mind, it all comes down to the Office.NET test. That is, at some point in the future (if MS sticks with .NET), Microsoft will release Office.NET (the Office suite rewritten for .NET). The test is: will Office.NET run, off the shelf and without any major crashes or missing functionality, on a non-MS .NET framework.
.NET as the greatest cross platform programming environment in the universe and hold the developers of the non-MS .NET frameworks up as champions of the open source world.
.NET frameworks as a waste of the open source community's time. And while, of course, programmers are free to waste their time if they want, don't expect me to be interested and contribute.
.NET apologists might claim that the Office.NET test is unfair. That there are many useful things that can be done with a partial implementation of .NET. To that I say, you're deluding yourself. If the best that we can hope for is a two .NET world (MS .NET and the open source .NET ghetto) then, as a professional programmer, it's not something I'm going to use. My time is important enough that I'm not going to waste it trying to remember what bits of .NET I can and can't use when I use Mono or dotGNU.
.NET framework. They'll keep back just enough to insure that their Office cash cow continues.
If it will, then I'll be happy to declare
But if it doesn't, then I judge the whole effort to produce non-MS
Now,
Now, as you can probably tell from the tone of my post, I've more-or-less concluded that there is basically no chance that MS will ever allow Office.NET to run on any non-MS
I think this is the big advantage for Mono. I am a GNOME zealot, but was not excited about mono at all. But after seeing the ease of doing GNOME/GTK work in Mono I have to say it looks very interresting.
I'm seeing alot of people complaing about how mono is a waste of time because MS can pull the carpet out from under everyone with proprietary library updates that will prevent .NET apps from running on mono. That is entirely probable and it is most likely beyond the ability of the mono team to keep up with such changes. Let alone the time to ensure .NET to mono compatability. But give the mono team some credit, don't you think they too might know this? I ask this question to all the waste of time whiners: What can MS do to stop mono to .NET compatability?
.NET, without any rewrites, and with a consistent feel. That is the point. It would speed the cross-platform development of projects like Mozilla and OpenOffice that both windows and linux users alike are trying. What better way to transition users dependent on windows apps(not windows itself) to a linux environment, one app at a time. Forget about allowing the commercial .NET programs to run under mono, think about replacing them with equivalent Open Source mono apps.
That's right, open source projects written in mono will have the ability to run under both linux and windows, mono and
I like the mono project but I get an icky feeling when I see .exe and .dll files. I don't know, they just make me feel icky.
Great quote! I love how it ignores basic facts of the players in question, not to mention the very, very long history of Microsoft's monopoly and abuse of the market.
Nope...sure wouldn't want basic facts known, for people to make a decision.
You can either look at history and nod knowingly, this is going to be bad for everyone except for Microsoft, or you can be a sucker and get kicked in the nuts, some time down the road, for the 1000th-time, and try your best to act suprised. While everyone around is laughing and pointing at the sheer stupidity of it all, I might add. Meanwhile, +10 points for Microsoft. Good thing you were there to help them. They sure needed help with their monopoly!
This is exactly what MS will do. How freaking naive can MS weeines be? MS is all about dominating a market. Big deal if C# the language is a standard. The platform is what matters and MS controls their .Net platform. People/Companies will code to that MS .Net platform and leave all other platforms out of the game.
If you want a cross-platform framework, use Java. If you want a cross-platform application, use wxWindows or QT
In the most recent Linux Magazie or Linux Journal (I don't recall which since I get both), I just read that many analysts are predicting that Linux will have at a minimum 45% of the server market share by 2007, only three years away. If Linux gets that kind of market share within 3 years, don't you think MS will continue to do whatever they can to continue to lock customers into their platform? Why do you think MS ported their C# compiler to FreeBSD over Linux?
Again, C# the language is open and you or anyone else can create a compiler. Big deal, you can now compile your first Hello World! C# applications. The commercial C# applications will be built on .Net and limited to MS Only.
Is there anyone out there that actaully thinks MS would develop a true cross-platform solution? Can their be a human that is that naive?
If Tyranny and Oppression come to this land,
it will be in the guise of fighting a foreign enemy. -James Madison
It seems Mono made a mistake in relying too heavily on Wine. It may have been more work, and may have been a duplication of effort, but they would have been better off in the long run recreating what they needed to, especially if their goal is 100% compatibility. Thus, they can only attain that 100% if (and its a big IF) Wine also attains it.
Manipulate the moderator system! Mod someone as "overrated" today.
I work on the Portable.net project
.net gui exe compiled in Windows, copy it to a Linux machine and run it and it works!
http://www.dotgnu.org/.
Ive written a large chunk of the System.Drawing and System.Windows.Forms namespace. We currently have two "toolkits" that means our SWF copy will work on Windows and X Windows (using Xlib directly). Mac OS is thus supported.
C++ improved on straight C, Java improved on C++, C# improved on Java. C# is becoming an important standard.
We want to reuse software. We want to take existing software that is built to only run on Windows and run it on Linux or other platforms.
We have much of the framework built to write gui applications using System.Windows.Forms. Despite peoples initial suspicions, what we have done works. We already have most controls completed, including textboxes, treeviews, comboboxes, tabcontrols etc.
See http://pnet.homelinux.org for some screenshots
You can now take a
The technology is excellent and open source. We need people to help us finish it off.
I understand where you're coming from, but none of these points pertain to the .NET Framework as a Framework, but rather what you can build with it.
.NET Framework, and GTK# is Mono's extension to the .NET Framework to allow more portable .NET graphics.
.NET standard doesn't, but which could be added to .NET. They're extensions.
.NET.
.NET application. If you wrote a reimplementation of .NET that had more capabilities in its APIs than Microsoft's, then you'd be extending .NET just the same as Mono is.
No. You can't build GTK# using only the Framework, any more than you could build Windows.Forms using only the Framework. You need GTK for GTK#, and you need Win32 for Windows.Forms.
Windows.Forms is part of Microsoft's
The Framework is the CLR and standardized API.
Right. And Mono implements the CLR decently (minus a few features), then implements the API decently (minus a lot of features, but slowly catching up); then it adds to that API more features.
Constructing their own GTK# language
API, not language.
and database access is not an extension to "The Framework" but rather an implementation of the "The Framework".
Wow. No. Not even. Ever. Under any definition.
Both of those things (GTK# and the database API whose name I've forgotten) are APIs that Mono developers have, that the
They are NOT a implementation of the Framework. Mono is an implementation of the Framework.
Unless they are making modifications to the CLR and/or adding new instructions to the CIL, they're not extending anything,
Great! Then we agree -- because Mono makes modifications to the CLR (by providing more APIs), it's extending
[they're not extending anything,] anymore than I do when I write my own application.
But Mono isn't just a
-Billy
Um, ya, that was kind of the point.
Let me draw you a chart. If Mono can acheive 100% compatibility with the code THEY are responsible for, but there are compatibility issues with Wine, than, overall, there is no 100%. Since they have no control or influence over the Wine project, they will have, at that point, reached an impasse; since Wine can improve, not improve, fold, or accept external improvements to their sources and summarily wipe their asses with them, there is nothing further Mono can do to improve their own project.
If, however, they impliment their own windows emulation, they can make use of changes in the Wine project (giving them credit, of course), start from scratch, or whatever. But, they then have a situation in which they can acheive their goals, and have the control they require to further their main project.
As people higher in this thread have stated, there are issues with Wine. Were this not the case, such a statement would be senseless. But all duplication of effort is not necessarily a bad thing.
If you ever learn how to manage a project, this whole subject is basic.
Manipulate the moderator system! Mod someone as "overrated" today.
The question is, will these developers avoid .Net if there was no Mono? If not, then a few years from now you'll have the killer apps ported to the .Net API, and alternative OSes still won't be able to run any of them. If Mono is able to achieve significant compatibility, then a port becomes that much cheaper to make. Maybe even cheap enough to make a profit from Linux sales.
The commercial C# applications will be built on .Net and limited to MS Only.
MacOS X and Linux can probably claim at least 10% of the desktop market. MacOS X currently can attract major applications like PhotoShop. What if you could support twice the MacOS X market, using the same code base as your main Windows product, by avoiding a small set of Microsoft-only APIs?
The commercial C# applications will support Linux if enough Linux users are willing to pay enough for them to justify the port. Today is port is very expensive. Tomorrow there will be more Linux users, and Mono may lower the porting cost significantly.
Mono & dotNet appear to be good technology, but that's already been discussed enough.
So...
With regard to Mono, Novell must either:
(a) have done a complete legal analysis of what will happen when MS doesn't like Mono anymore, or
(b) believe that Mono will always be acceptable to MS.
The fact that they haven't told us about (a) makes me fear that the truth is (b).
Novell, if you are listening, please tell me the answer. I'm a developer and I like the dotNet technology, but I need to know where you are going with this, and I need to now whether it is 'safe' for me (and my conscience) to use Mono and, for example, your windows forms library.
Windows.Forms is part of Microsoft's .NET Framework, and GTK# is Mono's extension to the .NET Framework to allow more portable .NET graphics.
.NET framework. Gtk# is a completely different library and toolkit from .NET. Gtk# is a C# language binding of the Gtk+ library. You don't need to know .NET in order to use Gtk#.
.NET just the same as Mono is.
.NET compatibility. But if that subproject were dropped from Mono, you'd still be left with a vibrant, useful project. It is therefore wrong to talk of Mono as "extending .NET" as if .NET was the foundation of everything Mono does.
Gtk# is not an extension to the
then you'd be extending
Mono is a project that comprises many subprojects. One of those subprojects is