Updated: 4-Jul-2009 To add OIN patents and applications.

Why Mono is desirable

That’s a question which depends on whom is being asked.

If you’re asking a user, then the answer is “it’s not”, any more than a Scheme compiler or LOLCODE interpreter.

Mono provides a well-balanced framework to enable Free Software to be developed quickly, effectively, and efficiently. By “balanced”, I mean it is unlikely to win every single contest that one can throw at a programming language – memory footprint, execution speed, availability of libraries, and so on – but does fairly well in ALL of them. Speedwise, Mono is much faster than Python – up to several hundred times faster according to some benchmarks. It has a fraction of the memory footprint of Java applications. It has modern features such as garbage collection which make it easy peasy to write code with compared to malloc()-happy C or C++. It is a well-balanced framework. As such, for people looking to write apps for a Free Software environment, it offers a compelling choice of framework on which to build. Several apps which have only existed for a short time – such as GNOME Do – make full use of functionality provided by Mono in order to be written very quickly and easily, compared to chasing SIGSEGV around due to human errors inevitable with C-based development. Mono was first conceived as a way to escape from the absolute horror of maintaining a large GUI C codebase (Evolution).

Alright, the problems starts here:

Well-balanced framework

This is no great argument for or against mono – most languages have strengths and weaknesses, and it is up to the developer / project manager / corporate monkey to chose what language(s) to use. Being “well-balanced” is something most languages / frameworks strive for, trading off one area for another according to whatever set of criteria deemed relevant. There are a host of “well-balanced” frameworks, not the least of which is Java.

Mono v. Python

Comparing Mono and python on speed is ridiculous – you might as well compare Java and Logo. Who exactly is making the argument that Python (interpreted) is faster than C# (compiled)? So here we see the favorite target of mono apologists: the straw man. Knock down something no one ever said and pretend like you just made some point.

Mono v. Java

Sure, sometimes mono has a fraction of the memory footprint. Other times, it just might be the other way around:

How about we compare like-to-like?

So, sometimes Java wins, sometimes mono wins. Sometimes “wins” depends on whether you want to trade memory use for speed. To simply state “mono has a fraction of the memory footprint of Java applications” doesn’t tell us much.

C++ is “malloc” happy

Uh, no. If you are using malloc in C++ you are doing it wrong. But hey, don’t listen to me, listen to the man himself:

[malloc, calloc, free, realloc] should be avoided in favor of new, delete, and standard containers.

- Bjarne Stroustrup, The C++ Programming Language, 19.4.6

Any dozen of books on C++ will tell you the same. I know because I have them on a shelf behind me. Now, either Mr. Shields knows this and is being dishonest by misrepresenting C++, or he is too ignorant to even speak to the most basic aspects of the language. I will leave that judgment up to the Gentle Reader.

Gnome Do

So Gnome Do and “several” apps which have only existed for a short time use mono. Great. There are a dozen more in a dozen other languages as well. Pick any comparable language and you’ll find the same story. This isn’t an argument for or against mono – because good coders can knock out good apps in any given framework. The implication is that only mono can offer such results, which is obviously not true.

Horror of large GUI C codebase

Finally, we are told of why mono was conceived, as if that were some sort of argument in its favor! It doesn’t matter why it was conceived, only what it actually is. Of course, the intent here is to imply that mono is a success in this area, which may or may not be true – we can’t tell from the excellent facts laid before us.

Moving on ….

Taking it further, Mono on the whole also enables easier migration – for both developers and users – from legacy CLR frameworks such as Microsoft.NET. Students who learn Visual Studio.NET at University can take their skills and directly apply them to creating or improving Free Software on their shiny new Ubuntu installations, without the need to learn a new language. Businesses with investment in .NET-based applications can look at replacing their servers or desktops with Free Software. Whilst providing .NET compatibility has always been a secondary goal, it is an extremely popular one, which has prompted a lot of input and development work from assorted people into the Mono codebase.

It should be noted, however, that this use-case (Windows migration) is not cause for including Mono by default (any more than, say, Wine), and indeed, the libraries required to run the majority of Microsoft.NET applications are excluded from Ubuntu installations due to lack of necessity.

Here’s another “big lie” of the whole mono thing: that mono is somehow enabling hordes of Windows developers to start developing for Linux. I can tell you what I see: lots of mono development on Windows, iPhone, and other non-Linux platforms. Primary-Linux developers have never been shy about tackling cross-platform development. Primary-Windows developers, not so much. There are a couple of reasons for this:

• Primary-Windows developers just plain don’t give two shits about non-Windows platforms
• To do interesting Windows development, you need a lot of framework that is outside the scope of mono

So, yeah – not buying what you are selling. Tell you what, mono has been around for about 9 years. How about a short list of Primary-Windows apps that have come to Linux thanks to mono? Not the other way around – I already know Team Mono is knocking themselves out showing off MonoDevelop and other mono apps in Windows.

And of course, the follow up is that this doesn’t even matter by his own admission. So why even bring it up? Because at a first reading, it really seems like it is saying something meaningful.

Right. So that’s the entire first section and not one compelling argument in the bunch. Shall we continue?

Why it is not a threat

Mono is not the result of any deals between Novell and Microsoft. Mono was started 4 years before that unfortunate deal took place. Mono gets no special treatment under that deal. It is not mentioned in the deal (as with other apps). This is important to note.

Except it is mentioned. Right in the “Joint letter to the Open Source Community From Novell and Microsoft“:

Mono, OpenOffice and Samba

• Under the patent agreement, customers will receive coverage for Mono, Samba, and OpenOffice as well as .NET and Windows Server.
• All of these technologies will be improved upon during the 5 years of the agreement and there are some limits on the coverage that would be provided for future technologies added to these offerings.
• The collaboration framework we have put in place allows us to work on complex subjects such as this where intellectual property and innovation are important parts of the conversation.

Novell customers can use these technologies, secure in the knowledge that Microsoft and Novell are working together to offer the best possible joint solution.

So there’s that.

Mono is covered by the OIN, as with most other major Free apps. Patent attacks against Mono carry the same risk to attackers as attacks against any other OIN entrant. Attacks against Mono would risk patent “world war”, which Microsoft cannot win. Such an action would harm their business – and lose them money.

This is actually the first of substance in the whole piece. Of course, OIN didn’t say squat for Tom Tom, nor have they ever published the list of covered applications (which was supposed to happen back in February 2006), nor have they ever done anything, but it is true that 3 years ago a dude from Red Hat said mono was in there. And, of course, Microsoft could attack through a proxy, ala SCO, or maybe even feel the fight was worth it one day. So while I consider this somewhat re-assuring and really the only worthwhile argument made in the whole piece -  it’s no slam dunk.

Sometime since I last checked the OIN site they have actually listed applications. Mono version 1.1.10 is right there on the list. So, as long as you are using Mono 1.1.10 and the OIN has the goods, money and will to go head-to-head with Microsoft, then this point may actually comfort you. By the way, the OIN lists nearly 100 patents owned. Microsoft has over 5000, perhaps over 10,000.

Mono implements an international standard – albeit one from a convicted monopolist. If this is a problem, why do people use C, the standard from convicted monopolist AT&T? Mono implements an improved, Free replacement for a proprietary offering. If this is a problem, why do people use GNU (which provided an improved, Free replacement for proprietary UNIX)?

Sigh. Because AT&T was a convicted software monopolist that attacked the FLOSS community for two decades? And AT&T’s was in direct competition with Linux and FLOSS? Because just last year AT&T was fined $135 Billion dollars for non-compliance with past anti-trust rulings? Oh … AT&T and Microsoft aren’t really all that much alike after all? But that would mean someone is trying to be dishonest in the argument! Say it ain’t so!

Shit man, why not bring up Standard Oil or the damn East India Company? Same relevance.

And, of course, no one has a problem with an improved, free replacement for a proprietary offering. Indeed, from one point of view, the entire point of Free Software is to provide an improved, free replacement for proprietary offerings. Of course, the implication is that mono is an improved, free replacement – which is not quite so clear cut. It’s certainly questionable that it is improved – it does have some unique namespaces. However, it also just as certainly lacks many features that “real” .NET provides. As for free, I think part of the debate is what the true cost of mono is.

Regardless of whether or not any specific patent licenses over ECMA 334 and 335 cover Mono’s implementation of those standards, if indeed such agreements are available (ITWire’s curlish “attempt” to secure such an arrangement aside), the fact that statements have been made in public supporting the idea of royalty-free licensing essentially reduces the financial impact of such infringement to zero. If Foocorp has a license to use patents, under a “non-discriminatory” license, and did not pay for them – then it would be discriminatory to change anyone  else for them (breaking the signed terms regarding patent licensing), and as such, those patents lose any financial value. They may, however, still hold non-financial value (such as their use in defending against patent-related attacks), hence not making the patents “free for all” in any understood sense.

This is wrong on so many levels, I’m not sure where to start, but … How about this: mono – right now, inarguably – implements namespaces that are not part of the ECMA standard:

Microsoft: “Dammit judge, we did all we could. We even submitted the core parts to international standards, and those dirty communists just aren’t content with that! Shut ‘em down!”

Judge: “Dirty Hackers! Give up that code! Find in favor of the prosecution.”

Or how about this:

Microsoft: “Judge, all we are asking for is a simple public acknowledgement on each source file that the code contains Microsoft patented IP. That’s reasonable and non-discrimnatory, don’t you think?”

Judge: “Dirty Hackers! Give up that credit! Find in favor of the prosecution.”

Or how about this:

Novell and Microsoft already got together to find a pin-hole sized loophole in the GPLv2 and rode the biggest wagon train of internet feces flinging man has ever known right through that tiny opening. This is why neither of them can ever be trusted to stick to what you or I might think is a reasonable interpretation of the spirit of any agreement or license. They have already both shown if they think it needs to be done they will find a way to make it happen.

You have to assume the worst because the parties involved have shown they will do the worst.

Mono cannot be “disabled” via an incompatible change to Microsoft.NET, for two reasons. Firstly, such a change would also break every single existing app for Microsoft.NET (in fact, if it were to happen, then the best option for those users would be to run their apps with Mono instead). Secondly, support for Microsoft.NET is a secondary goal for Mono – if Microsoft change something in .NET 5.0, then so what? It doesn’t prevent Banshee or GNOME Do from compiling and running fine on non-legacy systems like Ubuntu.

Except Microsoft can very well provide a clean upgrade or compatibility path if needed a dozen different ways, and if they “break” away totally, then who the hell is going to use mono? The whole damn point is to be a .NET clone. If you aren’t cloning it any longer, you have no reason to exist.

Absence of “patent protection” is not the same thing as “patent violation”. If I offer to sell somebody a promise not to sue them using any of my patents, their taking me up on my offer is no guarantee that I even have anything valid to sell them – merely that they are willing to buy it. If Jim buys protection from any patents I hold, it does not mean that Jim is infringing on anything specific – nor that if Ted does the same thing, that Ted is violating anything either. If a house insurance policy includes flood protection, it does not mean that your house will be flooded – and not buying flood protection does not mean that you will be flooded either.

I guess this is true as far as it goes, however: Novell and Microsoft continually throw around code words like “piece of mind” and “respecting IP rights” which all play into Microsoft’s FUD and patent aggression. Microsoft doesn’t have to have a valid patent case, they just have to convince enough suits they do. Novell is helping them do that. I hope no one is trying to argue that the Novell covenant appears to weaken Microsoft’s position that Linux infringes patents, or that Microsoft has portrayed different in public.

Patents covering a specific implementation detail of a project cannot kill it dead – the example here is Freetype. Apple made some patent threats against the Freetype developers, for using their proprietary hinting data stored in TrueType fonts. You’ll notice that Freetype still exists today – this is because the specific METHOD that apple laid claim to was worked around, and auto-generated hinting data used instead. Apple’s threat was diffused, and the project went on. Suggestions have been made by FUD vendors that the Linux kernel contains a number of patent infringements – if specific details are ever provided, then those specific infringements can be worked around. A patent infringed in the Linux kernel would not cause all GNU/Linux distributions to be shut down overnight – and it’s disingenuous to suggest that any other Free Software is any different on that front. Even if a fundamental patent is infringed in Free Software, it’s no big deal – as a core change can be made, and applications etc which expect the “old” way can be modified to work with the “new” way, easily. Only proprietary applications cannot be fixed in the event of radical change – and proprietary apps are a secondary concern.

The whole difference here, of course, is that no other project is working hand-in-hand with Microsoft, using exclusive testing packages from Microsoft and other vendors, nor are other projects trying to position themselves as the foundation for Linux development. Be honest here: the mono project is aiming to be the primary choice for all Linux development. By de Icaza’s own admission he would like to see GNOME itself built on .NET. That greatly complicates “work arounds”.

The layering of escape routes is extensive in Mono, especially Mono in Debian/Ubuntu. In the first instance, the contentious Microsoft-sourced non-ISO libraries such as System.Windows.Forms are not included by default, and are rarely used in Free applications anyway (because WinForms looks like ass, amongst other things). If a reason is found to remove these non-standardised libraries, then bam, they’re gone – without harming Free apps. Secondly, if a more severe change is required, then the Mono packages can be patched to remove the infringement. If more drastic changes are required, as mentioned above, then the applications can also be patched to support any core changes. And, taking it one step further, if the whole of Mono needs to be pulled, then applications can be ported. The porting process would potentially be slow and painful, and cause great harm to an application’s future rate of development (although no worse than if the app was written in the target language from day one), but nobody who has written an app they care about would simply throw their hands in the air and say “never mind, it was fun, but the Man says I have to stop now”.

Except, of course, the mono project will continue to add non-standardized components. This is exactly as Microsoft plans: if one were to limit the project strictly to the ECMA-covered portions it would be near useless. So, in order for the project to be attractive it must continue to add more and more non-standardized components. I’m not making this up; here, in Microsoft’s own words:

We have to spend a lot of money to make sure the openness of C# is well understood and that it is accepted at a level that allows our innovations to have traction. [...] The strength of this platform and the innovation around it is the key element in preventing commodization by Linux, our installed base and Network Appliance vendors.

- Bill Gates, 2001 “Highly Confidential” Email

Maintaining Gap vs. Linux

1. Keep network effect with Applications

• Migrate applications to .NET framework
  • BUT keep framework proprietary to Windows
  • Patents required to implement clone

[...]

Maintaining Gap vs. Linux

• The .NET framework contains the latest developer platform for the future, and it must be licensed like Windows. Subsets have gone about as far as they should go in the standards bodies, but we need a compact subset for phones and TVs. It was noted that we have to be careful because once the horses are out, they are out forever. At the right royalty, we can have discussions around technology beyond this.

- Dan Nault, 2005 “Highly Confidential” presentation given to Bill Gates, Steve Ballmer, and others.

I would like to understand the x-plat strategy, because I don’t get it and nor does anyone in the field. It doesn’t seem to make ANY sense.

By putting CLI into ECMA, we are inviting x-plat implementations. With Rotor we are even doing some work on Linux and Solaris.

For enterprise customers/partners, wanting to build enterprise apps, all this is meaningless as there is no mechanism to provide transaction (and other core services) support on these non-Windows implementations. J2EE clearly has a mechanism, albeit faulty, to enable these x-plat services.

I can think of four explanations for our current strategy (as I and the field see it):

1. There is something happening to provide these applications services x-platform that I don’t understand (based on Joe [Long]’s proposition, I doubt it is this).

2. We think that our customers/partners/analysts are stupid and that they won’t see our approach as insufficient for real enterprise apps.

3. We are going to evangelize to IBM and others to plug their own transaction services under CLI on Linux (without specifying how). Joe’s proposal is to tell them how.

4. We haven’t thought thru a strategy that will make sense after anything more than a superficial inspection – if so why are we doing all the Rotor work?

If the answer is (2) then we have learned nothing from the past 5 years and J2EE will continue to kick our butts. I would rather see Microsoft say x-plat is BS rather than make a half step (Rotor, CLI) that will confuse everyone and lead to continued distrust of our motives for doing it.

-Graham Clark, GM .NET Business Development,  2001 “Confidential” Email

I could go on (and on) with the quotes, but I hope you see. The idea that Microsoft submitted C#/.NET to ECMA without fully considering what they were doing is the highest levels of naiveté.

Why Mono should be included in Ubuntu by default

Here he basically argues that some mono apps are “best of breed”. This isn’t really a productive area for discussion, because if you put the mono issue totally aside it comes down to what you like in an app. I think any two reasonable people can prefer different applications.

However, there are a couple of points to make. Note well that although Tomboy was trotted out as a big, great app there was no mention of Gnote. Gnote almost single-handedly puts the lie to this entire “best-of-breed” argument and remains under surprisingly vicious attack from mono apologists.

The second quick point to make is that mono apologists love to ride that slippery slope. Once they get one little toe-hold in, they use that as justification for shoveling in all sorts of mono apps. Furthermore, Moonlight is on much, much, weaker ground than mono – but it seems to be pretty much the same people pushing them both. This is problematic and shows a slippery slope in a different direction.

The post on Linux Today

Not much in here but some back and forth. A few nice ad hominems, especially calling mono critics “software terrorists”.  Stay classy, San Diego.

The anti-Mono “movement”

There is a bit of uncomfortable truth here actually. There are some anti-Mono people that are way out of line. It is never acceptable to go after people’s jobs or personal lives. Period.

That being said, there are some real jackasses on the pro-Mono side too.

Furthermore being said, the end is largely the same attempt to paint any mono critic as a crank or loon. This is a favorite tactic of poor debaters everywhere. I believe it is the “You are a poopy-head” logical fallacy.”

Mono critics – at least this one – are indeed “for” something. I am for Free Software. I am for a strong, united FLOSS community, free of the influence of Microsoft-as-it-currently-exists.

It is because I see mono as a net loss for the FLOSS community that I am against it. I see years already past of divisive strife in our community – strife that Novell knew would come out of the Microsoft covenant, but they proceeded anyway. (I’m intentionally glossing over the pre-Novell mono years, which are just as offensive but I’m tired now.)

I see years of continuing strife every time “mono” comes up because mono apologists can’t give solid answers to critics or questions. And it’s not totally your fault – I think Microsoft likes it just fine this way. It’s a pure win-win for them. They aren’t going to “put things to rest” because it’s just not in their interest to do so. And that in itself is another reason that shows Microsoft is not to be trusted. They could easily squash any problem I (or any reasonable person) has with mono is a half-dozen ways. You know it and I know it. But they won’t. Because they do not want GNU/Linux to succeed.