Miguel de Icaza defends Mono and talks about its future relationship with the Gnome desktop, in the latest LugRadio.
Miguel de Icaza on Mono
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
I recommend it highly…
I think that .NET is the best development platform in existence. But even if this were not the case, mono is very important.
The fact is that there are literally millions of windows programmers that know .NET. If the applications they produce can be easily ported to linux and osx, this is a great opportunity.
I agree with you Rüdiger.
It will be very important to Linux to have a full compatible .Net framework.
I can wait for Windows.Forms. And better than this will be Borland releasing a Delphi for Mono.
Yeap is greate to know more about the Mono project and project of Novell, very good talk.
“I can wait for Windows.Forms. And better than this will be Borland releasing a Delphi for Mono.”
I think better than that would be for Borland to maintain/revamp Kylix something they’ve abandoned since 2002.
“I think better than that would be for Borland to maintain/revamp Kylix”
I strongly agree. I really wish instead of scrapping it, they’d just asked those of us put off by it what we would want changed before plunking down some money. I loved delphi and was extreamly excited about kylix. But there’s not a chance I’d use it when the resulting programs added yet another look and feel to any system it was on. I’m betting a lot more of us would have jumped on if they’d just allowed it to pick up the current KDE theme, given that it was using QT anyway.
Aside from the Ximian team at Novell, where is this army of developers and flood of applications that are going to seamlessly port to Linux?
not sure the gnome community will use .net….
50% of the people want to use it.. and another 50% don’t want to use it?
i don’t think Miguel de Icaza will change something if the gnome community don’t want his toy
tough to see which side of the fence you’re on <sarcasm />
Is there a transcript available?
what’s wrong supporting mono… .Net is a great developping platform..a java fork yeah but something microsoft is pushing hard. What we have to do now is agreeing with the “technology” and do our best to support what is being used. We want people to use GNU/Linux yes or not ? Then a great step is to support microsoft technology because it correspond to more than 90% (a guess…) of the market.
if you’ve kept, up even a little, with the mono stuff you won’t learn anything here. The part with Miguel is a short part at the begining, and the rest of the dialog is only [very] mildly interesting. I didn’t quite finish listening yet but I don’t expect that they’ve saved the best for last.
Read the slashdot article. While there is plenty of miguel bashing I’ve noticed that there are a lot more defenders than ever before. It’s nice to see.
I like mono because I like how fast creative/mature apps are being provided that use it.
While I understand peoples concerns, to a point, I think most critics are somewhat ignorant of the overall framework (and the legal risks). In the end I think mono will be viewed as an important piece in Linux’s success.
I wish people would stop harping on mono and .net if they aren’t developers or if they haven’t really looked closely at .net. The one thing that I (and Miguel also I think) really likes about .Net is the ability for any language to be a Producer, Consumer, and Extender. This makes it far cheaper and easier to maintain software in the long run and use existing software modules in a new or different language. This also makes it far easier and cheaper to come up with new languages that stand a chance at being accepted, which I would think some of you would care about even though the name of the site isn’t Compiler News or Language News.
why don’t you add a couple of non-question comments so we can all go “why, why”.
Yes. This is quite a lot different from bindings or just 1 or 2 languages being able to operate well together. I think there are already about 80 some languages that target .NET. I’m sure you’ll want me to post a comprehensive list … perhaps I’ll work on setting one up sometime.
on windows majority of the people will use ms .net not mono…
Yes, it is true. But the windows platform far more important for software developer companies then linux, because the 95% of the users are use windows, at least on the desktop.
mono is not 100% with the ms .net…
Yes, but IMHO in long term it can run many windows applications (like wine under linux) and the port of the .NET based applications can be easier, like Delphi/Kylix.
java is a better choice in this case, more mature, faster, more software engineering tool…
If you have a professional software developer company, the safe future more important then multiplatform develpment. The future of java is a little bit ambigous. IMHO the SUN is no too stable company if compare to Microsoft. IBM also support java, but on the other side also support .NET. And there are open source projects like klasspath, gcj, etc, but IMHO it is not enought for a bigger company.
> IBM also support java, but on the other side also support .NET
IBM absolutely does not support .Net in any form or function as it would be completely counter productive to their predominantly Java based strategy. It is going to be a cold day in hell before IBM starts throwing resources behind .Net. They have already got extremely good thing going with Java. If Mono is to seek any support from commercial vendors out there, it is going to be a very lone cowboy indeed because there is only Novell that is sort of interested in Mono, the rest of the crowd is seriously disgusted with it (even M$ is not interested).
> Yes. This is quite a lot different from bindings or just 1 or 2 languages being able to operate well together. I think there are already about 80 some languages that target .NET.
Well, I am quite confident that there are more languages running on JVM than on .NET. Check http://www.robert-tolksdorf.de/vmlanguages.html
Since Robert’s vmlanguages list is a little uneven in qualities, I will list what I consider to be quality JVM language: Java (of course), Jython (is Python), JavaScript (with Rhino), BeanShell, Groovy (in JSP), Kawa (high performance Scheme), JScheme (another Scheme, with emphasis on Java integration), Jacl (is Tcl), Nice (static typing, compiled, but with type inference).
Microsoft has indicated that if Mono becomes a problem they won’t hesitate to destroy it through legel matters.
This is BS. The stability crap is e-mailed to people who talk to IBM about sun. I e-mailed IBM and I had someone else e-mail them. they said the same thing. They hint that Sun is changing direction and is not very profitable, being unstable. They push this hard on sun and some people actually believe it!!
I even saw something about being more multi-platform use .Net??? .Net eats up more resources than Java. Microsoft has .Net not only for a good archetecture but so you can run platforms across the different CPU’s that windows runs on (AMD64, IA-64, x86, ARM).
Face it. Microsoft wants a .Net-Lock in. Liking them because the are a big company is nonsense. I remember when .Net first came about.. I’m surpised of the little adoption. It was supposed to sworm in and pile up but groth has slowed. Java is a better brand while .Net has too much confusion.
Oh, IBM has enterprise Java software that competes with Microsoft’s .Net.
Java is growing my friends.. ever so strong. I personally think it’s more flexible. OK! Well, it’s in BIG demand here and .Net is not in as big of a demand.
ummm, what? IBM does have *actual* products that target .net. Do a search on their site. XDE comes to mind but there are others for sure. And don’t you worry – Microsoft loves to tell people about mono – it validates their position when it suits them. “Novell is sort of interested in Mono”??? Are you ok?
> Microsoft has indicated that if Mono becomes a problem they won’t hesitate to destroy it through legel matters.
Care to back this up ?
correct me if I’m wrong but java wasn’t designed with this in mind while .Net was.
“Well, I am quite confident that there are more languages running on JVM than on .NET. Check http://www.robert-tolksdorf.de/vmlanguages.html“
correct me if I’m wrong but java wasn’t designed with this in mind while .Net was.
“”Yes. This is quite a lot different from bindings or just 1 or 2 languages being able to operate well together. I think there are already about 80 some languages that target .NET.”
How well maintained?”
Here’s a list of many languages that work with .net
http://www.dotnetpowered.com/languages.aspx
Please come back with a full repot of how maintained they are.
> ummm, what? IBM does have *actual* products that target .net.
IBM does not even a single .Net based product in the product line. There are some compatibility and interop layers added for .Net in products like DB2, but .Net is not taken seriously by any stretch of imagination by IBM.
“Microsoft has indicated that if Mono becomes a problem they won’t hesitate to destroy it through legel matters.”
Where, when, etc? This is news to me and I’ve followed pretty closely… Linkage please.
I wish developers would stop writing bloatware and start coding some quality apps in assembly.
Go back to roots.
Stop beomg lazy, stop the bloat!
How many freaking layers of code do you need to produce an app?
“Are they? Why would you say that? ”
Don’t get me wrong. There are a lot of people that know more than me about .net but I get so tired of the same stuff coming up. While some people claim there are still issues with the ECMA stack (CLI and C#) the chances of seeing this portion targeted are as unlikely as anything else out there (that would really make MS look bad to standardize something only to try and take it back). Still, more often than not, people still lump all of mono and .net together and call it a big legal risk. Regardless of what happens with the more proprietery portions (which will most likely be a non issue), mono will remain a great linux development platform (using the gnome libraries with the ECMA stack). This, more than anything, is misunderstood it seems.
If Novell really wanted to push Mono they should develop or support oficially a decent IDE for it, one that can at least compete with VS.Net. How the heck someone who uses VS.net will be convinced to move to Linux and use Monodevelop instead? Not in a gazillion years.
> How the heck someone who uses VS.net will be convinced to move to Linux and use Monodevelop instead? Not in a gazillion years.
Maybe not, but thats a personal choice … However Mono has got me to start writing apps for Mac OSX.
However Mono has got me to start writing apps for Mac OSX
Agree. I can confirm too that we are closely monitoring (as a company) MONO. When it will get better and more compatible, it will be fun to have such good platform to develop (hopefully) cross-platform (web)applications.
Moreover, I don’t think why people care to troll about this. If you like, then use it. If you don’t like, then don’t use it. It’s that simple 😉
However, we must say that at least Ximian/Novell IS delivering. Someone else pointed out that. And MONO is opensource (not that I care that much but for the ones who do…).
However, I agree with Moose too that tools are important. I wrote about this in another thread. I’m surprised that Novell didn’t put more workforce into developing a good IDE and that is leaving this task (as Miguel himself stated here) to the “community”. At least, they could have engineered some Eclipse plug-in or kind of that.
However, bashing is not useful. Time will tell if MONO will be successful or not. There’s no need to forecast that. For what concerns myself, I can only say my company (a Windows-only, right now) is considering a good thing to have a compatibility layer.
I only have one question: why there is no MONO version for Solaris? 😀
I’ve played around with mono a little bit and I’m very impressed. I have a background in C and Python and it seems to be at least as solid as Python, and I hear there is a Python environment that sits on .NET. The demos with Mozilla and Glade are very impressive — a full graphical web browser with just a little bit of code. Given some more time, I might move from python to mono for some personal projects.
> Here’s a list of many languages that work with .net
> http://www.dotnetpowered.com/languages.aspx
> Please come back with a full repot of how maintained they are.
Here’s my take:
Mature: C#, VB.NET, MC++, J#, all from Microsoft. I consider mcs(C#) mature. mbas is not there yet. Mono equivalent of MC++ is not started. But IKVM.NET is cool beyond any descriptions!
Note that VB.NET is not really same VB. It is VB changed to fit .NET restrictions. (This casts doubt on whether .NET really is language neutral, but see below.) Same for MC++ and J#. (Not as bad as VB 6 vs VB.NET, from what I heard.)
Compared to Java side, C# doesn’t run on JVM yet. (No equivalent of J# or IKVM.NET.) C2J compiles C to JVM and is mature. (It could compile all of PGP, for instance.)
Bridge, not really .NET language: RDNZL is a Lisp library providing access to .NET libraries — that is, a bridge. It’s one way as far as I know — .NET has no access to Lisp. It’s also AllegroCL-specific. Same for Python.NET hosted on Zope. It’s just CPython compiled to managed code. DotScheme too. It’s a PLT-scheme library. PHP Mono Extension also is a bridge.
Failure: Perl for .NET and Python for .NET from ActiveState was an utter failure, from what I could read from their website, and also from my experience using Python for .NET.
IronPython is a wonder, but far behind Jython for now. It needs time to mature.
Lua.NET is two part: one is simply a bridge, the other is Lua2IL compiler which looks promising but still very alpha.
As you can read by following the link, TickleSharp is P/Invoke wrapped Tcl, not really managed code. Compare this to pure Java Jacl. (Jacl has JNI companion TclJava too.)
Spry is a vyporware, from what I can tell. MixNet? You know, MIX is from Knuth’s book… Toy, not practical.
IL assembly: JVM has assemblers too. Jasmin and Jamaica to name a few. Also BCEL(ByteCode Engineering Library).
Interesting new languages: Nemerle: Compiled, agile features, type inference, macros. Comparable to Java’s Nice. Mondrian: functional, scripting language. Can be used in ASP.NET, Haskell-like features. Quite cool. Boo: they like to call it “Python-inspired”, but quiet different beast. Duck typing. Maybe like Groovy.
Some serious and very cool stuffs: S# and F#. Efficient(!) Smalltalk and OCaml implementation. This may invalidate my doubt about .NET language neutrality. But I also heard that S# and F# is essentially a big hack, working around .NET’s idea of what programming language should be, to death. SML.NET belongs here too, but I don’t know much about it.
What’s so good about Cw(Comega)? I don’t understand what are its advantanges at all.
Summary:
There are four .NET languages: C#, something-like-VB, C++ with non-standard extensions, not-really-Java. .NET also has IL, official text assembly format for bytecodes, while Java doesn’t.
Smalltalk and OCaml implementation (S# and F#) may be impressive in showing .NET’s language neutrality. Nemerle, Mondrian, Boo are nice new languages.
There are many bindings, bridges to .NET platform. These are *not* .NET languages. There are attempts to port existing languages like Python and Lua, but these are preliminary.
Does anybody know what the hell is going on with that? I check the mailinglist every once in a while and it’s dead. I know the creator got hired by Microsoft (a great thing). I think Miguel had stated that they would be willing to host the project if need be.
C# is a nice language (a step up from Java IMO), but like Java it’s really a systems language and not so RAD.
I think the mono website should really put more emphasis on the difference between mono and .NET
It is not a .NET implementation, it is something similar and aims at good compatability.
Currently I see this as a problem, lots of people, even developers, even more decision makers, get the impression that Mono is a .NET implementation, thus allows deploying of applications of .NET
While this may be almost true now, there is no guarantee that it stays that way.
The history of MS Java has shown that MS is very likely going to ship extension libraries with their base VMs and SDKs and that a lot of developers will not be cautious enough to see the differences.
So they will still consider their applications portable, something I would consider mostly relevant for server-side portions, but will in fact bind themselves to Windows servers.
When they find out, it will be too late, a lot of money would have to be spent to remove all those extensions again.
I guess it is really helpful for marketing mono when you blurr the difference between it and .NET, but it is IMHO a very short sighted advantage
Yeah I am also disappointed with IronPython’s lack of activity. Maybe the guys as Redmond hired Jim Hugunin just to strap him to a chair?
At least Boo, by the other hand, is very active (at least from I could see in the mailing lists) and a nice community around it is showing up. They even built a plugin for Sharpdevelop (with autocompletion and everything), and managed to create ASP.Net pages with Boo as the code-behind language..not bad.
For those interested, the site is http://boo.codehaus.org
Yeah, I’ve checked out Boo. It’s nice – type inferencing and all, and has a good idea where it just leverages all of the .NET framework for its libraries, but I’ll wait for it to mature a bit before thinking about serious development.
The guys that were working on Prothon or Proton (forget the name), which was a prototyped based language with Python-like syntax are working on another language that rides on top of .NET/Mono.
Maybe not, but thats a personal choice … However Mono has got me to start writing apps for Mac OSX.
You’re not serious, are you? What Mono apps are available on OS X? Gtk# isn’t available for OS X yet via official channels (doesn’t ship with Mono for OS X for some reason), is a pain to compile via fink, and still uses X11 and looks like a major sore thumb on OS X and doesn’t have all the necessary key bindings for it to [u]behave[/u] like a native app. This has been the biggest criticism of Swing apps on many platforms. At the moment, Gtk# is better than Swing on X11; Frankly, what isn’t better than Swing on X11? I don’t know how it fares on Windows, but on OS X it is far from being a good development choice.
I’m glad that Mono has made you think about cross-platform development. But there are many better tools for the job, like Qt.
ummm, what? IBM does have *actual* products that target .net. Do a search on their site. XDE comes to mind
XDE is a product they acquired when they bought rational.
How about giving a little work to port Cocoa# to GNUstep?
And since GTK+ is being integrated with Cairo, then it may be
easier to fit a GNUstep theme into GTK+ than fitting GTK+
into Cocoa. I guess that it is easier to port GNUstep
to any environment than GDK too imho, hence, less effort to
maintain and improve things.
…is a pain to compile via fink, and still uses X11 and looks like a major sore thumb on OS X and doesn’t have all the necessary key bindings for it to [u]behave[/u] like a native app. This has been the biggest…
And again, as soon as someone mentions the availability of a non-native OS X app, we immediately have an Apfel user clamoring about non perfect looks, non perfect keybindings, non perfect something. So every time someone actually mentions the portability and availability of Free Software on OS X, we have other three Apfels cursing and threatening not going to use any of it at all, for all time.
miguel is a nice guy, lugradio now needs to interview pygtk guys [miguel referes to those guys and says: “not enough books”]. Last but not least, the interviewers need to grow up
Java does not offer anything along the lines of module reuse in the same way that .NET does:
“A particularly important aspect of compliance is the CLS level. CLS stands for Common Language System; it’s the specification of three subsets of the Virtual Object System, adherence to which will ensure full language interoperability. The subsets are:
* Compliant producer: this ensures that your components (by avoiding non-universal mechanisms) can be used by anyone.
* Consumer: if your compiler satisfies this, your classes can reuse, as clients, components written in any compliant-producer language.
* Extender: if your compiler satisfies this, your classes can extend classes from any compliant producer language, that is to say, inherit from these classes and redefine (override) their operations.”
This is taken from Bertrand Meyer’s article at http://archive.eiffel.com/doc/manuals/technology/bmarticles/sd/dotn…
Its a great read if you really want an informed third party opinion on the pros and cons of .NET
Anyone here actually use Win32 API or MFC to write code for windows, NOW?
I do. I can’t wait for .NET–I hate the current state of the win32 API AND ALL OF THE DIFFERENT VERSIONS OF IT. Just using bitblt() sucks.
About how compatible is Mono with .NET? What exactly isn’t implemented and what exactly IS NOT GOING TO BE IMPLEMENTED. Is MONO + associated tools suitable as a replacement for Visual Studio.NET ?
” The one thing that I (and Miguel also I think) really likes about .Net is the ability for any language to be a Producer, Consumer, and Extender. This makes it far cheaper and easier to maintain software in the long run and use existing software modules in a new or different language. ”
that really a big joke
approximately thirty percent of the projects are a success… a success means to finish the project in the costs, in times as well as the system meets the user need
imagine now if in a project, many programming language are used?
You’ve obviously never used any X11 apps on OS X. But the problems brought up are valid. It isn’t a problem with F/OSS software so there’s no need to get defensive. Loads of OSS software that are written in Qt work fine on OS X with little modifications, and they behave natively.
If you have used X11 apps on OS X, you will see that the menubars are in the wrong place, the key bindings are all wrong, font rendering is generally poor, seriously, just find a mac, and launch some X11 app on it. The problems aren’t merely aesthetic in nature. It severely hampers productivity especially if you’ve got to memorize two sets of keybindings for two types of apps (Native and X11), two menu bar locations, and struggle to read stuff on screen.
It isn’t only the apple fan(atic)s who complain about the UI of their apps. You wouldn’t tolerate a strange looking/behaving app on any DE. Unless you were absolutely forced to use it due to lack of options, you will go for the most productive system.
Some F/OSS software works good on the Mac. Octave and LyX come quickly to mind. I’m sure there are others. There’s no need to get defensive when people bring up real issues with ported F/OSS software.
Thanks for the heads up. Didn’t know about Cocoa#. How compatible is Gtk# and Cocoa#? By that I mean are any changes to code required to port an existing Gtk# app to Cocoa#? The documentation on that page seems quite incomplete and there doesn’t look like there’s a way to download Cocoa#.
Look at cocoa#.
http://www.mono-project.com/using/cocoa-sharp.html
It’s still at a very early stage of development, but should address issues people have with non-native X11 apps on OS X.
I guess this will make mono a valuable tool for OS X, or cross-platform, development.
I guess i missed the prevoius post about cocoa#.
GTK# and Cocoa# are different GUI toolkits. So you would have to rewrite the GUI part of you app to port it to OS X.
But that is a design choice. For your app to be 100% native to a specific OS, you should write a sepperate GUI layer for each (GTK#/Linux, Cocoa#/OS X, Windows.Forms/Win).
And having the GUI layer seppeare from the bussines logic is almost always a good thing.
“imagine now if in a project, many programming language are used?”
too bad that you haven’t picked up 3rd party tools and software to save money and to deliver on time – then you would know how nice it is to continue writing in c# while some of your imported softwares are in vb .net or whatever language it happens to be in.
Basically, you should be able to achieve cross-platform portability if you use Windows.Forms.
I think that GTK# and COCOA# will be extensions but the standard .NET GUI set should be enough to achieve portability. Of course, Windows.Forms will use GTK# to rendere GUI on a Linux system while it will use COCOA# to do that on OS X (as well as it will use native .NET objects to render on a Windows system).
I believe that you don’t need to directly support GTK# or COCOA# unless you need to write platform-specific code. Anyway, I didn’t do any test about this… I’m just speaking according to what I understand from info.
Digging through various open sourced .NET libraries, and tools I’ve discovered that most of them are both .NET and Mono compatible. Look at NUnit, NAnt, Log4NET or Gentle.NET. And there are others. So, MONO already has plenty of tools just because it appeared early enough to let .net programmers to care about it. Contrary to Java, where most important OSS software, like JBoss, Hibernate, Eclipse or most of Apache stuff etc. is troublesome to run with gcj/Classpath or Kaffe (to put it mildly…).
“too bad that you haven’t picked up 3rd party tools and software to save money and to deliver on time ”
we deliver on time, we use software engineering pratice
“then you would know how nice it is to continue writing in c# while some of your imported softwares are in vb .net or whatever language it happens to be in.”
don’t need to use that toy, the market use c++ or java
Back before I learned Python I used to be very concerned about the whole Java/Mono/.Net thing. Now I’m not so concerned.
“Note that VB.NET is not really same VB. It is VB changed to fit .NET restrictions.”
You make it sound as if VB .Net is LESS capable in .net. VB has gone through an upgrade because many people complained about functionalities not being there.
“Compared to Java side, C# doesn’t run on JVM yet. (No equivalent of J# or IKVM.NET.)”
dotGNU claims that this works.
“Failure: Perl for .NET and Python for .NET from ActiveState was an utter failure, from what I could read from their website,”
They still sell it from their website.
“There are four .NET languages: C#, something-like-VB, C++ with non-standard extensions, not-really-Java…”
it’s still vb (even with your doubt), c++ is actually Visual C++ which is a Microsoft product but was also touted as one of the most compliant versions of c++ available. j# – a no-brainer I suppose. You left out JScript .Net, but it’s ok, everyone does.
“we deliver on time, we use software engineering pratice ”
right, and all those .net idiots don’t.
“don’t need to use that toy, the market use c++ or java”
So are you saying if the market decided, you would use c# over java or c++?
“And what did programmers do before C# came along then?”
port, wrap, bridge, etc… I guess.
I’m talking about writing in vb .net, inheriting in c#, and instantiating in c++ – all this works out of the box (if written to be cls compliant).
” right, and all those .net idiots don’t. ”
the .net tool for take metric, testing… are poor and not really numerous
when java go out, a lot of engineering tool go out too
” So are you saying if the market decided, you would use c# over java or c++?”
people learn what the market want
there are a couple of year, majority of the university courses was done in c++, now it’s java
when the student go out, it will use java
it’s company who decided
“the .net tool for take metric, testing… are poor and not really numerous”
So now it’s tools and not “software engineering pratice”? Like Miguel said in the interview “It’s ok, you don’t have to like .net” (or something very similar).
Uhm… This was reported 1 week ago or so, why has it been put around 10 articles up as it used to be ?
“So are you saying if the market decided, you would use c# over java or c++?”
i work on an enterprise ERP/Production workflow management j2ee webapp. who do you think decided what technology to use? not me, thats for sure. it was management, who decided to go with the industry standard. we also have a mammoth of a vb6 app which is more targetted to sourcing stuff that needs to be upgraded. after looking at the pros and cons, our company decided to rewrite the whole thing in java rather then upgrade to .net. for those of you who dont know, the word “rewrite” tends to make managements heads explode, they loathe work that doesnt give (visible) added value to the customer, and have only a vague idea of why its important. for them to give the go ahead for a (pretty much) total rewrite of the app over using the ms migration tools just goes to show how much trust they have in .net as a viable enterprise platform. java is a proven standard, .net is a newbie put out by a company with a horrendous track record on things that enterprise apps require (i.e. stability and security).
to sum it up, it will take an act of god to get .net seriously accepted in the enterprise market anytime soon. asp has been around forever now, and j2ee is still almost universally considered the way to go for serious enterprise webapps. where you are going to see .net flourish is in desktop apps, not enterprise apps (which is what a substancial part of .net is aimed at). the j2ee/nix/oracle stack is pretty much dominant atm, and that isnt going to change any time soon.
“to sum it up, it will take an act of god to get .net seriously accepted in the enterprise market anytime soon.”
I think you should go talk to Mark Andriessen about “acts of god”. For good or bad, it happens.
“that really a big joke
approximately thirty percent of the projects are a success… a success means to finish the project in the costs, in times as well as the system meets the user need
imagine now if in a project, many programming language are used?”
I never stated many programming languages should be used during the primary development of a project, although with .Net it would indeed matter less. I’m talking about people, for instance, that have huge code bases in something like cobol. I personally know developers who are quite good at getting the job done in cobol but aren’t really that great at anything else. Sad but true. Something like NetCobol (http://www.adtools.com/info/whitepaper/net.html) would allow these same developers to make use of the many libraries available to .Net, as well as making porting legacy code over to .Net easier.
“to sum it up, it will take an act of god to get .net seriously accepted in the enterprise market anytime soon.”
hahaha … well if we always go with the tried and true, then we will never come across anything better. I’m not saying management was wrong to pick the java platform since it is proven to work quite well and .Net shares many of (but also has many in addition to) the qualities of Java.
I see jobs offers all the time that want people with .NET experience, in fact, I see these a lot more than ones looking for Java experience.
“I see jobs offers all the time that want people with .NET experience, in fact, I see these a lot more than ones looking for Java experience.”
go on any big job search like monster and count job for both
java is most used, it’s a fact
more enterprise want to go out of ms monopolization, that begin by not used .net
anyway, with the money invest in java, enterprise will not stop there development with that technology
I’m talking about people, for instance, that have huge code bases in something like cobol. I personally know developers who are quite good at getting the job done in cobol but aren’t really that great at anything else. Sad but true. Something like NetCobol (http://www.adtools.com/info/whitepaper/net.html) would allow these same developers to make use of the many libraries available to .Net, as well as making porting legacy code over to .Net easier.
Exactly! I know some too. Imagine how many banks still have Cobol codebases and tons of stuff. Being able to leverage upon such knowledge and code while you’re on top of a modern framework could save you big bucks.
A friend also gave me a book about Cobol.NET (which I never read because I don’t know Cobol, of course 😉 but this comment is definitely about one of they key feature in .NET. You might like, maybe you don’t but for many scenarios that could be an huge difference.
Plus list of supported or partially-supported languages for CLR is huge and impressive. Inheriting from class written in C#, extending in VB.NET or J#, overloading in Cobol.NET or F# and so on is a very powerful way to mix knowledge. Think how many times someone you know didn’t get a job because he’s experienced in VB while company requires Java, or experienced in C++ while company requires VB.
.NET allows company to focus on knowledge rather than other details. So if you are good at VB.NET, damn-develop that rocking class in VB and forget any concern: other developers will simply import it, inherit, extend, subclass or just use it.
Think what you want but that looks an advantage to me 😉
it depends what kind of jobs we are talking about. the microsoft platform is great for small/medium sized businesses. but if you are writing for fortune 500 companies, chances are you are doing it in java. the very fact you think enterprise businesses are going to use anything but the tried and true shows you dont know too much about the field.
youre right, it does happen. and i have no doubt .net will be a huge success, its just i dont see it biting into javas market any time soon. its just too new, and its by microsoft. those two things will make descision makers very wary before jumping on board. when .net is a mature technology, with an api that rivals javas, that is when things will get interesting. but we are still a few years from that.
To both Chris and brandon. Don’t either one of you realize that a lot of languages allowed you to call other languages even before .NET. It’s not like people before .NET were tied to using one language. They did for other reasons. e.g. training, convienence.
“Exactly! I know some too. Imagine how many banks still have Cobol codebases and tons of stuff. Being able to leverage upon such knowledge and code while you’re on top of a modern framework could save you big bucks.
”
Imagine how many don’t have the source code to begin with, and therefore CLR and all that is irrelevent.
Your argument seems self defeating; at one point, Java was in the same place that you assert .NET is in now. Granted, Java is already accepted and is MUCH better than the collection of technologies being used beforehand. Still, if .NET is used long enough by smaller business, larger businesses will catch on and start using it slowly. And that’s not to say it hasn’t started to happen already: http://www.microsoft.com/net/momentum/
I admit I don’t have the industry experience you do, as I’ve never worked for anything approaching a Top500 business, and I respect your views for that reason. Its just that your particular quote saying that .NET will never be accepted in the Enterprise made me laugh.
Can you give an example of a framework or platform that offers language interop to the same degree as .NET? All I have to do is create an assembly in one .NET language and reference it from another application.
Also, what do you mean the source isn’t available? I would hope that its available to the company that wrote it …
I missed your post directed at chris, so I see that we generally agree with each other. Personally I’m looking forward to Mono maturing, but as can be exemplified by this lug radio episode, people do hate MS tech – even when they do something right.
ya – I think I can recall something like that from some part of my brain. That and I do see benefit to *not* using more than one language for a project too. The ease of interop doesn’t suck I have to say. Either way, this isn’t even my favorite feature of .net/mono/dotgnu etc… my ADD must be kicking in because I forget which part of this discussion was interesting 24 hours later …probably started with all the mono bashing
miguel the icaza my ass! –: )))))))).