OSNews

Unlike most .NET supporters, I don't think .NET is the greatest thing since sliced bread. It is a pretty good thing though. Just because Microsoft comes up with something doesn't instantly make it "BAD". .NET brings an entire set of easy to use and fully capable APIs along with countless different languages can use those APIs to create OS agnostic (but perhaps limited to x86 and x64) applications. Sure .NET is slower than native code in most cases and sure it uses more resources than native code in most places. But the rapid application development it allows for while simultaneously supporting such diversity has not really been achieved with much else.

From a corporate standpoint .NET is a fantastic platform for in house development. Development time is crucial so the RAD is a huge plus. And with the advent of Mono most .NET applications can be easily deployed to Windows and Linux servers both 32 and 64 bit all with a single build.

Silverlight / Moonlight simply brings this same environment to a browser.

While there are countless situations where .NET is simply not an appropriate choice, the same can be said for virtually any other software development technology.

No, coming from Microsoft doesn't make it bad from a technical point. It makes it bad because Microsoft is openly campaigning against the Linux platform. There is NO patent protection in using their platform. FOSS is better off steering clear of this technology. Adobe on the other hand, is opening the specs on their products so that GPL'd versions (Gnash and others) will actually work right without having to reverse engineer everything. There are also opening up their development tools to Linux. When Microsoft actually starts working with open standards and file formats, then it will be time to consider their technology.

There is NO patent protection in using their platform.

If I understand it correctly (and I very well may not), the patent protection was one of the big things in the Novell-Microsoft agreement. It was essentially an agreement of patent and lawsuit protection between both parties. Perhaps I could be wrong but that was my understanding of it after reading it.

Edited 2008-05-15 15:18 UTC

So that means you are tied to Novell/Microsoft not just Microsoft.

Edit: And I'm not sure exactly how much the Novell conenant not to sue actually covers.

Edited 2008-05-15 15:48 UTC

MS has opened the specs of the .Net CLR, C#, and Silverlight. In fact, MS has been helping the Mono guys do Moonlight, with docs, support, etc.

I get as mad as anyone at Microsoft for their business tactics, and some of their software being crap, and not being totally customer focused. But some software MS does is very good, and MS isn't always evil. So a little level headed balance is in order.

Always proceed with caution when dealing with MS tech compatibility. But don't reject it outright.

BTW - I just downloaded the Moonlight plugin into my Firefox, running on Mandriva 2008 Spring, restarted Firefox, then went to the MS Silverlight demo website, and it worked flawlessly.

I'm glad I have the choice to use this. There are, and will be, websites that use Silverlight, and I love that I now have the ability to view those sights on Linux.

Also, as a developer who works in both a Windows world and *nix world, I'm glad I have an sdk for .Net and Silverlight to run on Linux.

Also, back to the tired old concern about MS patents - the .Net CLR and C# are open ECMA standards with "no sue" declarations, and MS has completely opened up the Silverlight specs to the Moonlight guys, and even helped them with docs and technical support. Also, it is very much in Microsoft's interests to have this *nix implementation, and with no effort on their part.

So, in short, don't worry, be happy - but still keep a weary eye

We'll see how much support (and the Mac for that metter) they get if Silverlight becomes as popular and well-used as Microsoft hopes ;-).

No. What Microsoft does is partly open the specs. Some vital part it keeps for itself.

.Net is a classic example:
http://en.wikipedia.org/wiki/Microsoft_.NET#Standardization_and_lic...
"In August 2000, Microsoft, Hewlett-Packard, and Intel worked to standardize CLI and the C# programming language. By December 2001, both were ratified ECMA standards (ECMA 335 and ECMA 334). ISO followed in April 2003.

While Microsoft and their partners hold patents for the CLI and C#, ECMA and ISO require that all patents essential to implementation be made available under 'reasonable and non-discriminatory (RAND) terms'. In addition to meeting these terms, the companies have agreed to make the patents available royalty-free.

However, this does not apply for the part of the .NET Framework which is not covered by the ECMA/ISO standard, which includes Windows Forms, ADO.NET, and ASP.NET. Patents that Microsoft holds in these areas may deter non-Microsoft implementations of the full framework."

See? You can have most of .Net ... but you can't write a GUI because Windows Forms is held back. If you try to port Windows forms to another platform, so that a program originally written in .Net using Windows forms can be easily ported to that other platform ... Microsoft would certainly sue.

Absolutely classic Microsoft behaviour, that.

For Silverlight ... the piece of that that Microsoft have held back for themselves exclusively is the content creation part.

You can only create Silverlight content on a Windows platform.

Ergo, it is totally unsuited to become a standard ... just like most of Microsoft's stuff is similarly unsuited.

Are you saying that Microsoft is spreading patent FUD, or that people are spreading FUD about Microsoft spreading patent FUD?


Microsoft are the company who are the most detrimental to open source software, have described it as 'communist' and have done everything that they can reasonably do to deride it in every way possible.

It's really quite funny, and sad, that all that people have in response to that is that people have some sort of anti-Microsoft agenda and are Microsoft haters. The same thing was said in response to every objection over OOXML. Oh, it's all an IBM backed conspiracy and everyone hates us!

The ball's in Microsoft's court and they can give us a call.

Red herring. Microsoft has made public guarantees that it will indemnify customers from patent issues (as long as you're not suing them, yourself) when using its technologies.

As long as they get it under an agreement from Novell ;-). They've made no such guarantees for anyone else or open source developers.

It's a game you've got to ask yourself if it's worth playing, or whether you even need to.

Well heres an example of your guarantee. In an article (Microsoft fights piracy through openness) on Inquirer.net, the author talked about Microsoft's new push to be open.

"When working with Microsoft's patented communication protocols, the company will provide a promise or covenant to open-source developers so that they could do their work without further needing to get their patent rights from the company."

In response to this statement, Microsoft senior vice president and general counsel Brad Smith was quoted as saying:

"But then if open-source software is distributed commercially, for example by a company or used commercially by a company, then we would expect people to think about our patent rights. And if they need a patent license, they could come and get one from us."


So what? We only get left alone if we don't make any money or try to distribute it? Thats some promise. I guess Microsoft has a patent on making money now too. Only someone who is completely clueless would trust Microsoft not to sue them if there is money involved. And how the hell does a promise stand up in court? Its not a legally binding contract under law.

You have heard of Java, right?

http://www.robert-tolksdorf.de/vmlanguages.html

If you don't have enough time to read the list of over 200 different languages, Jython, JRuby, OCamlJava all exist (JRuby in particular is much farther along than IronRuby).

I too thought the JVM was only Java, until fairly recently when after OpenJDK 6 was announced I actually took the time to look around.

I think the lesson is that branding is important, and up until fairly recently the branding/marketing was all Java. But technically speaking, that's never been the case.

Edited 2008-05-15 17:55 UTC

You best look up the number of languages supported by the JVM... Not only are "common" scripting languages like python and ruby supported, but also "new" languages like groovy and scala are making headway.

Sounds like Java, what are the major differences between both?

So that's why Tracker, a project that started out much later than Beagle with a lot less funding has more features, is more stable, consumes less memory and indexes faster? Given the so called Mono flagship applications of Beagle, F-Spot and Muine, how many people actually prefer them to their natively written counterparts?

People who criticize GTK+ have a point. However, when given the chance users will never ever tolerate an application that runs in a VM (even if there is a JIT). Such applications start up slower, take more memory and frequently run slower than their native counterparts.

It would be better to come up with GTK+ v3 or something and try to bring GTK+ inline with the clean Qt4 API.

I haven't seen Silverlight or Moonlight but if Moonlight will work on a 64bit browser, then the answer is no..flash is not any better, it is much worse.
Adobe did finally got Flash to use ALSA, so now they're only 5 years behind.

On a side note, someone told me that I can use some himem kernel option and actually use all 4GB of RAM in a 32bit Linux. This is looking pretty attractive to me.

Moonlight will work on 64 bit browsers, some of our developers develop on Linux x86-64

If you have PAE support enabled in kernel (most modern distros do) and your CPU also supports PAE (again, most modern CPU's do), you will be able to use up to 64 GB of RAM on a 32bit Linux.

PAE is also available for Windows but in desktop versions (XP, Vista) it only allows access to 4 GB tops. To get more you need a server version.

http://en.wikipedia.org/wiki/Physical_Address_Extension

Silverlight/Moonlight have been getting a lot of "shit" from FOSS and such lateley... however, is it all that worse than Flash?

Yes! Because it's developed by a company with a very strong interest to protect the monopoly position of their own operating system.

Silverlight/Moonlight has the word TRAP written all over it. Avoid!!

Edited 2008-05-15 17:34 UTC

Just like Adobe tries to keep its flash monopoly.
Just like the FSF tries to keep its GNU Tools monopoly.

Those two are not better than MS in that aspect.

Just like Adobe tries to keep its flash monopoly.
Unlike MS, Adobe has no reason to make it work better on one particular operating system. MS can stop releasing new versions of Silverlight for Mac when it has got enough market share. And for Linux Moonlight will be one step behind ("almost compatible").


Just like the FSF tries to keep its GNU Tools monopoly.
What are you talking about?

Edited 2008-05-15 17:56 UTC

Adobe is somewhat less dangerous than MS in that their dominance is somewhat easier to topple.

The FSF is hardly a monopoly. They are a software provider with tools which simply work exceedingly well, and have a sometimes annoying leader, and some overly vocal fans who do more harm than good.

I use market share and "barrier to entry" criteria to gauge monopoly status. But I set the gauge threshold to be the price/functionality ratio needed to cut into the dominant player's market share. That can place the barrier at a relatively high absolute level. FSF has good stuff. But the balance may well shift in the future.

Nobody wants any monopoly, but the fact is that Microsoft controls an OS and the stack all the way up to Silverlight. At this point in time that is infinitely more of a problem than something that runs on at least a variety of platforms today, and where Adobe has an incentive to see that it stays that way. Microsoft simply doesn't, and they will drop Moonlight and the Mac version of Silverlight like a hot potato if Silverlight achieves what they feel is critical mass.

They have no interest or incentive in non-Windows platforms, and you have to be so much of an idiot it isn't even funny if you still don't understand this. Do you really need this explaining to you after all these years?

As for Adobe, one hopes that they will come into line with HTML5 and keep producing tools which people buy anyway and where they will have an incentive to do so. The fact that they don't control an entire OS, development platform and browser means that this is much more likely.


Come again? This is open source software (or as RMS calls it, free software) which means that people can pick up development independently and fork from any FSF or GNU influences should they so wish. Many have already done so.

What the hell is your definition of a monopoly?

I was wondering if I could try it out on Safari.
Thanks for the info!

My 2 cents: In a multiOS open source world, where users of all kinds blend together, is it really that smart to make your application Linux-only by ties with ALSA or what else? I mean... isn't this what we are running away from in the first place?

Crap? Taking away potential political bias, the .NET platform seems pretty OK to me. Sure, a virtual machine is may be slower (or arguably faster in some situations), the .NET framework makes it largely possible to use the same class library and platform with different languages. Additionally, C# is a nice language. It seems to be largely inspired by Java, and there seems to be some cross-pollination now.

Sure, there is the potential of litigation, but probably not over ECMA-submitted parts. And Mono buys good code compatibility to what seems to become one of the standard platforms on Windows. For some odd reason people seem to bash Mono, but have no problem against the inclusion of WINE or Gnash in free *NIX systems. Double standards?

I think .NET provides healthy competition to its free software counterpart, Java. Mono brings compatibility. I can't see what is wrong with that, as long as no critical components or one-of-a-kind killer application depends on Mono, no ships are burnt.

One could say that compatibility with such technologies make it more popular. That could be the case for OS X, but I think with the small desktop usage of Linux or BSD, there is not so much influence over these technologies. Supporting them makes Unix-like systems more attractive to newcomers. I have seen many friends who were interested in GNU/Linux (primarily through Ubuntu), go away because of the bad support for some media formats. The average user out there wants to watch missed TV broadcasts, not C code .

Edited 2008-05-15 14:48 UTC

The zero cost RAND agreement that covers the ECMA specifications (and which Microsoft are bound to provide as ECMA members - not the zero cost bit by the way) is exceptionally flimsy, doesn't stop Microsoft having or acquiring patents and can easily be revoked by Microsoft getting the standard cancelled.


WINE and Gnash aren't covered by the above, where something looks open but really it isn't. Mono would have been better off ignoring the ECMA specifications all together and coming up with an independent implementation like WINE or Gnash.

With WINE, it just happens to be a piece of software that happens to be moderately Win32 API compatible. The same applies to Gnash. With the ECMA specifications what Microsoft gets is a way to tie down different implementations and say "Ah, that is covered by this!"

It's fairly clever really.

Beyond that, I hope Mono will be at least moderately useful in providing at least some compatibility as WINE has done. As with all open source software, someone goes out and implements it and we see what happens,


Microsoft isn't going to help you there. Windows Media is not covered in Moonlight at all ;-).

Edited 2008-05-15 21:45 UTC

While I do groan a bit as we enter another "let's struggle to be compatible with Microsoft" phase, you are correct. It is necessary. I am rather hoping that Mono, much as I am averse to it, will be somewhat more successful in its efforts at MS compatibility than was Wine.

*We* did?

You only bother to show up when your products are featured here, Miguel.

Edited 2008-05-15 23:43 UTC

I didn't ;-).

The above is the situation I'm afraid Miguel. The RAND (and zero cost) licensing is only covered for as long as the specifications are ECMA standards. Microsoft can rock the boat to get it revoked at any time, and the only recourse the ECMA has is to cancel the standard. Microsoft probably won't do this and will just stop contributing, but the thing is, they can. The ECMA provide you with no protection at all in these matters. You implement at your own risk. The ECMA also doesn't stop patents being held now or acquired in the future on their standards, just that they can be licensed accordingly. Read the FAQ on the ECMA's web site regarding this issue. I've looked everywhere to try and find that this isn't the case, as I got piqued by it in the past, but it is.

http://www.ecma-international.org/memento/codeofconduct.htm
http://www.ecma-international.org/memento/guidance.htm

This is why there was some commotion some while back, as I understand it, to try and guarantee that the RAND and zero cost licensing is irrevocable. No such guarantee has been forthcoming.

I hate patents and all this stuff as much as the next person, and I think you should just be able to ignore them, but is this really a web worth bothering with for open source developers? You do see what Microsoft is doing here don't you? They're trying to make the murky issue of whether you're covered by a software patent more certain where what they see as their 'intellectual property' is concerned.

Good luck.

Mono is an independent implementation. It is a ground up, clean room writing of the CLR and C#, based on the ECMA specs.

Gnash is also a clean room implementation of published specs.

Wine, AFAIK, is both a clean room implementation of published specs/docs, but also reverse engineering.

Adobe has patents on Flash, and could sue on them any time they wish.

MS has patents on Win32, and could sue on them any time they wish.

So, legally speaking, the risk level between Mono on one side, and Gnash and Wine on the other, is identical.

That said, I do trust Adobe to play nice more than MS.


Beyond that, I hope Mono will be at least moderately useful in providing at least some compatibility as WINE has done. As with all open source software, someone goes out and implements it and we see what happens,

[/q]

You're missing the point. It's an independent implementation of a specification that makes it CLR compliant. All of the patents you will see from Microsoft make reference to the fact that if you have implemented a compliant CLR via the ECMA specifications then you are covered. If you were to implement similar concepts in a JVM, for example, you are explicitly excluded.

It is not clean room in the strictest sense, because you're following a published set of documentation that is RAND licensed.


Not quite. It's very difficult to really tie a piece of software down to a patent that can be applied to it. With stuff like the ECMA RAND licensing and the Open Specification Promise, Microsoft is trying to solve this 'problem'.

I'm not really a Mono fan. But I am glad that they are independent. I don't exactly "want" Mono in Unix, but we need to have it just in case. Where would we be without Samba?

I was unaware that Moonlight was Linux only. That is unfortunate, but I don't imagine that it will be a problem for long. Ultimately, OSS is looking like the right choice for all of us.

I'm not really a Mono fan. But I am glad that they are independent. I don't exactly "want" Mono in Unix, but we need to have it just in case. Where would we be without Samba?

I was unaware that Moonlight was Linux only. That is unfortunate, but I don't imagine that it will be a problem for long. Ultimately, OSS is looking like the right choice for all of us. "

Well, it is Linux-only in that most of us use Linux, but this is open source code, so feel free to send us patches to improve the support for your favorite OS, or architecture, or browser.

In the last few days, some of our engineers landed support for Opera and WebKit on Linux, which is pretty cool. And some people on the #moonlight irc channel are porting it to Win32 (so they can embed the engine in their own apps).

So in the best spirit of open source, if you want Moonlight running on a different beast, use a different engine, get the source code and start sending patches.

Miguel.

I would like the choice of using this "crap" or not, stop trying to limit my choices. I choose the software I use on many factors, and usefulness, stability and speed come before manufacturer, or politics.

Haha, ya know the FOSS crowd is all about Freedom, until you attempt to use/build something they don't like.

Of course, Silverlight is bad for no other reason than the source code isn't publically available and it's made by Microsoft. So if you use it or any open source implementation of it, you will most certainly burn in hell, so sayith the Church of FOSS.

FOSS advocates are not all of a piece. Please keep that in mind. This issue is complex. I'm pretty steadfastly FOSS. And I think we need to keep our options open. That means having Moonlight at the ready, even if I do have qualms about it.

I agree, best tool for the job, that's what I say.

No. Silverlight is bad because Microsoft have reserved for Windows only the ability ... and more importantly the legal rights ... for the content creation part.

Therefore, it is not cross-platform.

Therefore, it is not suitable for use in any cross-platform context, such as ... the Internet.

The source code is out, we would be happy to integrate patches that would make it run with your favorite audio solution on your favorite Unix variant.

Miguel

What's this "nobody" crap? Thankfully, you don't speak for Unix. Why are you so afraid of additional choice?

If Netflix on-demand movies will run on moonlight/linux, then I care about it.

Yep, just like WINE and Samba and FAT support. And Java up until a year ago. And Flash up until last week. And DOC and XLS and PPT up until earlier this year.

I find it interesting that people enjoy being able to open binary Office formats and read FAT disks, and transfer files with Windows computers, and view Youtube, but wanting to run Windows programs on Linux is the greatest sin one can commit. (only via Mono/Moonlight of course, most people are fine with WINE.)

wine and samba are projects that were started at a much different time and for different reasons. you can't really compare the two (well, three). when wine was started, application support on *nix was not where it is today. and, ultimately, wine is a transitory tool. it helps people migrate AWAY from windows. as far as samba is concerned, simple networking in a mixed-platform environment is more urgent than a development platform of shady parentage. as far as binary formats and FAT are concerned, interoperability and supporting legacy devices and files are simply not comparable. nobody is using FAT or .doc because they simply love the tech. please... it's pretty clear that FAT and .doc are two things everybody is trying to get away from.

***.NET/Silverlight may be great tech.*** but a major reason so many FOSS supporters (and myself) have a problem with it is that it will always be a moving target, ultimately steered by a company that ultimately is not looking out for our best interests. you can scream that C# is an open standard all you want. that doesn't change who's steering the ship. that's not blind microsoft-hate. it's conflict avoidance, and there's nothing wrong with that. it's not irrational and it's not FUD. it's troubling that microsoft has a tendency to urgently need their fingers in EVERY pot. they openly want control of the "whole widget" and that bothers people. it may not bother YOU, but's it's a valid concern. they want to own the desktop, the server, search and advertising, console gaming. and it seems like they'll do almost anything to horn in on any market with potential. again, that may not bother YOU, but it's a valid concern.

and on a sidenote, i realize that people throw a lot of their hostility and negativity at miguel unjustly. (this thread is clearly no exception) but, frankly, it seriously appears to a huge number of FOSS supporters that everything he does revolves around emulating microsoft. so many people have worked so hard to *not need* microsoft (or their standards) at all. it's particularly unsettling to STILL be following their tails after all this time when, ultimately, it's COMPLETELY unneccessary. you don't have to agree with me, but acting like so many opinions are baseless is missing the point entirely.

I respect that opinion, and agree that a large part of Miguel's work does involve porting Microsoft technologies. Where we differ is that I think this is one of the most important things that can be done to help FOSS.

Despite the representation we may see on OSNews, Slashdot, or Digg, we still live in a world where ~95% of the world uses Windows. This is especially true in businesses where no one will upgrade anything because it might break something, much less switch to an entirely new operating system.

The ones that are willing to progress are still left with hundreds or thousands of internal custom applications that no one maintains any more. Rewriting these apps is out of the question. Miguel is interested in trying to give them options to move these things to Linux.

If you aren't in the 95% still stuck on Windows, you aren't the target audience. Just keep in mind that any effort that brings more people to FOSS helps all FOSS users have the numbers needed to demand things like better drivers and commercial applications.

Miguel may never win back the love of FOSS supporters, but I think he's ok with that. He is fishing in a bigger pond: the other 95%.

there _aren't_ hundreds of thousands of critical .NET apps that no one maintains any more. it's much too new for that. but, that's not really the point of mono. that is where WINE comes in handy, really. which, i don't believe miguel has anything to do with.



i have to respectfully disagree. the 95% still using windows generally *don't know a darn thing* about mono/moonlight. they are NOT the target audience. and, if they *were*, they don't care. the target of mono is developers. the target of moonlight is sorta users... but not many developers are using silverlight to begin with. and honestly, i hope nobody does. (see my previous comment) the last thing we should be doing is supporting a not-free-enough platform controlled by a not-trustworthy-enough company.

don't know why you got voted down, even though i don't agree with you it was a well reasoned comment

Here is the other side of the picture. This is a gap in technology that has existed since the mid eighties, and has been filled again and again by horrible solutions. the FLOSS community is either unwilling or unable to deal with web technology, it sucks, but its true. We have gotten a big fat nothing from them so far.

To be fair, its not like the commercial sector is that much better. I know everyone talks about them with awe and reverence, but W3C specs tend to be some of the poorest specs in existence. Overly complex, ambitious wording, and alot of just plain bad ideas that come out of consortium politics. It's no wonder nobody can implement this stuff properly.

Same goes for the rich application interface technologies. Java has been promising truly cross platform UIs for 30 years or so now, and have delivered poorly designed, unresponsive, and buggy results for just as long. Flash ended up gaining dominance, but it was never designed for what it is used for. As much as it is an incredible interactive vector animation tool, it is not fun to do serious coding on it.

So in 30 odd years that this has been a huge gap in the web space, where has the FLOSS world been? HTML 5 is still in whiteboard land, and to be honest, the technologies it uses are already obsolete. SVG has not exactly been winning any performance awards, and Theora is about four generations behind VP7 and VC-1.

I use technology based on its merit, If FLOSS is able to come up with a silverlight killer, I would be more then happy. But until then, the arguements dont really have legs to stand on. "Don't use commercial protocols, use nothing." is not a viable option.



That kind of fanatacism and backbiting is the #1 reason I no longer contribute to the linux community. There are some places I still hang out that have the feel of the old days, but I have no interesting in helping the community as a whole anymore.



This is what you don't understand. This whole thing was never about not using microsoft, it was about making great technology. That is what has always been the motivating force in the open source world.