Apache Declares War On Oracle Over Java

jfruhlinger writes "The Apache Software Foundation, feeling increasingly marginalized as Oracle asserts its control over the Java platform, is fighting back, trying to rally fellow members of the Java Community Process to block the next version of the language if Oracle doesn't make it available under an open license amenable to Apache. Last month's Oracle-IBM pact was a blow against the ASF, which had worked with IBM in the past, but it appears that Apache isn't giving up the fight."

Comments

Finally

[Yvan256]

Now we know who launched that missile!

Re:Finally

[Anonymous Coward]

Was it a tomahawk?

Re:Finally

[Lanteran]

Mission accomplished.

Re:Finally

[Arancaytar]

If it was launched by Apache, then it was probably a Tomahawk.

Unsurprising

[JustNilt]

Everything I know about Oracle makes this absolutely unsurprising. It looks to me as though they're trying to cut out all the "competition" in order to ride out the recession.

Re:Unsurprising

[mark72005]

There should be a top ten list of rising stars among evil companies.

(But who would hold slots #2-10?)

Re:Unsurprising

[Old97]

Huh? Oracle has been evil for most of its 30 year existence. If you've ever done business with them you'd have experienced it first hand. They'd have been worse than Microsoft if given the chance.

Re:Unsurprising

[corbettw]

No, they're using the recession as a convenient time to buy up or smash the competition so that when customers start buying again, they'll only buy from Oracle. From a business point of view, it's genius.

Re:Unsurprising

[Fluffeh]

From a business point of view, it's genius.

I wouldn't go so far as to say genius, but certainly say it is simple forward planning. It is the simple result of: Where are we now, where do we want to be financially, in the marketplace and from a client base point of view in 3 years, what is the gap between where we are and where we want to be - and finally, how do we cover that gap.

In this case, it is clearly simple. We want to be in a stronger market position and to achieve that, we need to earn a higher market penetration. To do this we need to either buy, discredit or discontinue our competitor products. We have the money available to make a lot of purchases as well as the current market position to be able to drive a very large product towards the goals that will benefit us most.

The move from Apache is clearly a salvo from a company who can perceive this change and doesn't like where it is going as it will clearly impact THEIR goals negatively. If they can make enough of a stink/problem/thorn about it, then Oracle will have to realign their own thinking/planning to plan a slightly different path that avoids this big thorn/problem or account for the fallout and accept it as part of the bigger solution.

Re:Unsurprising

[Anonymous Coward]

I worked with a guy back in the 90s who was one of Oracle's first employees. He used to tell me how horrible it was to work there and how much of a complete asshole Larry Ellison was. Fortunately for him, being in at the beginning gave him a head start and he ended up making millions writing O'Reilly books covering Oracle Database topics.

Re:Unsurprising

[jd]

Not sure about the recission part, but they're definitely out to kill the competition. I fully expect the battle to get extremely bloody. Apple's sacrifice of their own Java implementation might well have been under duress, given this development.

It might be a good idea at this point to start looking at other languages. Since D is supposed to be "C# done right", it might be a language worth investigating. All you'd need is a portable virtual machine for it and you've a rival to Java that is (supposedly) superior to Java structurally. Tcl/Tk, Perl, Python and Ruby are already highly portable - although Perl largely shot itself in the foot with Perl 6 and Python did some serious self-inflicted damage with Python 3. Both should recover - after all, Python had just as much of a problem moving to Python 2 from the original form. Regardless, clearly there are potential competitors to Java if they can be mobilized.

If one or more of these can be embedded into multiple browsers (IE, Firefox, Chrome, Safari and Opera would be the obvious set and cover almost the entire browser market), Java would face some serious competition - at least at the browser end. Java applications and servlets would depend on whether the Java ABI was covered by the patents. If the ABI (in and of itself) is not IP-protected, then it would be possible to write virtual machines that run entirely differently than "native Java" VMs but which support Java objects. Bring GCJ up to Java 7 and have a backend to GCC that supports a portable virtual machine. You then have something that will handle existing Java bytecode and will allow a gradual weaking off of Java to any language GCC supports.

(Since IBM -is- permitted to contribute to GCC, this is another direction IBM might be looking into. Especially if they can get a Java bytecode frontend working for GCC. Java applications natively compiled to IBM's processors would be very appealing, especially if it didn't break any standards in the process.)

Re:Unsurprising

[Hatta]

Haven't you heard? The recession is over.

Re:Unsurprising

[tarius8105]

When you buy oracle products just make sure you dont forget the $4.99 bottle of lube, it wont hurt as much.

Re:Unsurprising

[sinclair44]

I think/hope that they are going to absolutely shoot themselves in the foot with this. Much of their top talent has left in droves since the Sun acquisition. They sent a recruiting email to myself and some of my friends -- some of the top students at the top CS school in the country -- asking if we were interested in coming to work on the Solaris kernel full-time; they were pretty much collectively told, "After what you did to Sun? No way." If their talented engineers are by-and-large leaving and they are by-and-large unable to hire more, they will quickly become a dying shell of a mediocre company.

Re:Unsurprising

[keirre23hu]

Not sure about the recission part, but they're definitely out to kill the competition. I fully expect the battle to get extremely bloody. Apple's sacrifice of their own Java implementation might well have been under duress, given this development.

It might be a good idea at this point to start looking at other languages. Since D is supposed to be "C# done right", it might be a language worth investigating. All you'd need is a portable virtual machine for it and you've a rival to Java that is (supposedly) superior to Java structurally.

Yep and you better have a license from Oracle/Sun as Microsoft does or Larry will sick his patent police on you faster than you can say unsigned int

Re:Unsurprising

[Gazzonyx]

There should be a top ten list of rising stars among evil companies. (But who would hold slots #2-10?)

Oracle.

Re:Unsurprising

[Eskarel]

It's a bit more complicated that that really.

One of the reasons sun failed was that despite the fact that they came up with all the Java standards, the reference implementation and industry leader for most of them elsewhere. In this specific instance we're mostly talking about Tomcat as a servlet container which just destroyed anything Sun had until the very end. There was really no reason to pay either licensing or when they went open source support fees to Sun because their product implementations sucked.

Not that Oracle aren't a bunch of bastards(they are, always have been, and always will be), but Sun's relationship with the ASF was very much against their own interests, Oracle will very much be looking to put their implementation of the next JEE container as the go to standard so they can get some money out of it.

Re:Unsurprising

[Daniel Phillips]

One of the reasons sun failed was that despite the fact that they came up with all the Java standards, the reference implementation and industry leader for most of them elsewhere

That may be one reason, but don't forget that much of Java was grandiosely and stupidly overengineered to the point of suffering so much bloat as to make the platform undesirable for many or most possible uses. Besides Sun's inability to just let go and let the community fix the braindamage.

Re:Unsurprising

[illtud]

They sent a recruiting email to myself and some of my friends -- some of the top students at the top CS school in the country -- asking if we were interested in coming to work on the Solaris kernel full-time; they were pretty much collectively told, "After what you did to Sun? No way."

Unfortunately, I guess that your insightful feedback won't make it up the chain. All that Oracle HR will report is the number of new hires (and there will be some) that campaign made. They won't be top-class, on the whole (my opinon for the same reason you gave for flipping them the bird), and the Sun exodus will continue...

Please, if there's anybody out there who's considering sticking with Sun (ie the Sun products continued within Oracle) please speak up - I really need some pros to even make it worth while totting up the increasing number of cons.

Re:Unsurprising

[Eskarel]

Java isn't exactly bloated, most interpreted languages be they byte code or scripted require as much in the way of space and resources.

It's not exactly over engineered either, and a lot of what made it seem that way has been fixed. JEE no longer requires you to write an interface multiple times, and most of the stupidity with things like IO have been abstracted away(as they always should have been in the first place).

Java certainly has a few legacy issues from being the first byte code language and having made a few false assumptions, and J2EE was a monstrosity, but it's not actually all that bad in recent versions.

The biggest issue with Java is just that all the free VM's including Sun's suck.

Re:Unsurprising

[Daniel Phillips]

Java certainly has a few legacy issues from being the first byte code language...

Java was far from the first byte code language and has always suffered from having an inefficient byte code interpreter.

Come on, Java not bloated? Then why do the lights dim whenever I start up a Java application?

Re:Unsurprising

[mark72005]

I had a similar conversation with someone who worked for and with the co-founder of another fortune 500 company around the time the company started. I think most of these guys end up being slave drivers and assholes because they see what can happen and how much it's worth, and it just strips away all inhibitions.

Re:Unsurprising

[russotto]

JEE no longer requires you to write an interface multiple times

Wait, if you don't have to write an interface at least twice in code, then twice in XML configuration files and once in a .INI style configuration file, along with another XML file to put it all together, how do you know it's J2EE?

Re:Unsurprising

[Eskarel]

Well it's not J2EE anymore it's J6EE, but yes, that was pretty awful.

Re:Unsurprising

[Eskarel]

Because every single JVM I've ever worked with is a bloated monstrosity. The thing is that there's nothing fundamental to Java which requires the JVM to be that way, and for brief periods of time JVM's which didn't suck have existed. They just all seem to die.

Re:Unsurprising

[Anonymous Coward]

Right a bunch of students with no work experience unilaterally turn down guaranteed full time positions with a established company in this economy.

Re:Unsurprising

[mcvos]

I think the list of evil companies was usually headed by Microsoft or Sony. Or Exxon, Haliburton or BP, when we include non-tech companies. Oracle was always on the list, but not usually at the top of the list.

Clearly they're tired of just being the leader in Databases. They want to be the leader in Evil too. Much bigger growth market.

Re:Unsurprising

[CarbonShell]

I agree 100%
Although I come from the Java side of things, the feelings are mutual.
Already we are beginning to look for alternatives for Java based stuff.

As much as we all love [Solaris/Java/OpenOffice], Oracle is the one driving us away and that love won't stop us from leaving.

Google, rescue us!

Re:Unsurprising

[mcvos]

Java was far from the first byte code language and has always suffered from having an inefficient byte code interpreter.

Come on, Java not bloated? Then why do the lights dim whenever I start up a Java application?

Does your memory use that much power? Java is only inefficient if you're short on memory and have plenty of processing power. If it's the other way around, Java is pretty damn fast. Faster than C++ in some cases. There aren't a lot of languages, certainly not managed languages, that can claim the same. Well, Scala, obviously.

Re:Unsurprising

[CarbonShell]

'Let the community fix it'?
Are you kidding? The only thing the community would do is add in 6285 different variants of the same crap 'features'.

A community can do a lot, but they can also horribly fark up stuff.

I have seen some of the effects stuff like that can have in examples of C/C++/C# code.
It looks like you were reading it in a hex editor.

Also I think you might have to update your knowledge of Java. The criticism you seem to have looks like what people have been saying for the last 10+ years based on some early version of Java.

Re:Unsurprising

[TheRaven64]

If you've ever done business with them you'd have experienced it first hand

True, but there was a plus side. Unlike Microsoft, which employs a much lower grade of evil, it was very easy not to do business with Oracle. By the time you're a big enough customer for Oracle to notice you exist, you've got other big companies sending people around to try to persuade you not to use Oracle.

I don't really care if a company is evil if it's easy to avoid them. It's when a company becomes ubiquitous that it starts to be a problem.

Re:Unsurprising

[TheRaven64]

I wouldn't go so far as to say genius, but certainly say it is simple forward planning

They're the same in the business world. Why do companies like Apple, Microsoft, and Oracle do well? Because they have CEOs that are emotionally invested in the company and so expect to still be in the job 5-10 years from now. They make decisions that will make the company more profitable in the future, while a typical off-the-shelf CEO makes ones that screw the company long term but make a profit in the next quarter.

Of course, you also need some idea of strategy. Sun had two CEOs that cared about the company, but the first spent two much time grandstanding to notice he was driving it off a cliff and the second didn't have the business skill required to drag it back up again.

Re:Unsurprising

[TheRaven64]

Since D is supposed to be "C# done right", it might be a language worth investigating

Minor correction: it's supposed to be C++ done right. It predates C#: I first played with it in 2000 and it wasn't new then, C# was first released in 2001. It's not a bad language, if you like the Simula family.

Re:Unsurprising

[TheRaven64]

Why not, if they had other options? Given the choice between a job at Oracle or IBM, for example, I would be surprised if anyone took Oracle. If they're really at the top of their class, then they probably have a lot of options - I certainly did, although I chose to stay in academia for a bit and then work freelance so I never went through the whole 'proper job' thing.

I actually would have been quite interested in working on Solaris for Sun, but my contacts in the company put me off applying with their complaints about the poor working environment and bad management. I don't think I'd be interested in doing the same thing for Oracle. Mind you, one of my clients keeps talking about shipping a fork of OpenSolaris, so I might end up working on it after all...

Re:Unsurprising

[TheRaven64]

Java isn't exactly bloated, most interpreted languages be they byte code or scripted require as much in the way of space and resources.

This is true. The system requirements for Smalltalk were insane. The VM was written in Smalltalk, with a small subset that could be statically compiled and the rest interpreted as bytecode. To run a complete GUI windowing system and set of apps written in the language, you needed a 2MHz Alto with 512KB of RAM! Who could afford that? Fortunately, Brad Cox created Objective-C to give most of the flexibility of Smalltalk without quite so much overhead, so people could run it on machines that they could actually afford.

Re:Unsurprising

[Vectormatic]

MS is a clear case of "never ascribe to malice, that which can be adequately explained by incompetence". Balmer might give of some evil-vibes with the chair throwing and the occasional "Google, i'm going to fucking kill those guys", but in reality he is just a bumbling clumbsy guy trying to bring home the CEO bucks.

Larry Elison on the other hand.. i wouldnt be surprised at all if he has a secret underground lair under his house, stocked with plenty of black swiveling chairs, white cats and shark-pools (which may or may not be filled with seabass instead)

Re:Unsurprising

[FreekyGeek]

$4.99? If Oracle sold it, it wouldn't be a bottle, it would be a per-anus charge of $4,999 for each... "application".

Re:Unsurprising

[Civil_Disobedient]

Yeah, I agree. The GP smells trollish. Java has the benefit of being free and already widely-understood (hell, most CS intro classes start with Java because it gets OOP out of the way right from the start). If I'm hiring for a company, I want to have my pick of a hundred people, not three. D would have to give me magical performance increases in the range of an order-of-magnitude or more before I'd ever consider abandoning Java.

Re:Unsurprising

[mcvos]

MS has a longer history than just the bit with Balmer at the wheel. In the '80s and early '90s, they were very good at restrictive contracts, making everybody dependent on them and unable to get out. If they're not the masters of vendor lock-in anymore, it's because they've been hammered by the DoJ and EC.

I do think Balmer is evil, but he's too incompetent to be in the same league as Elison or the younger Bill Gates (from before he started to give away his money, and was still bent on world domination).

Re:Unsurprising

[maxume]

It's a little silly to accuse a billionaire of merely 'trying' to make CEO bucks.

Re:Unsurprising

[sinclair44]

It wasn't guaranteed, it was an interview offer (sorry that wasn't clear). We all do have work experience as interns at other companies, and most of us already have offers from places like Google, Facebook, Microsoft, and Dropbox.

Re:Unsurprising

[spiffmastercow]

No, that $4.99 bottle of lube isn't compatible, or at least not supported. Oracle recommends their own Oracle Intestinal Lubricant (OIL), which will cost you somewhere in the tens of thousands of dollars.

Re:Unsurprising

[tehcyder]

1. Oracle

2. Er, that's it.

Re:Unsurprising

[Eivind Eklund]

*Top* students have no problem getting full time offers. A lot of companies will extend them. It's the students that aren't stellar that can't afford to turn down offers; the top ones will have to turn some down, as they'll have more than one on the table.

Re:Unsurprising

[thetoadwarrior]

We use Oracle and I can't honestly anything nice about them. Their database isn't awful but all their tools are, imo, shit with SQL developer being on the top of the shit heap. Leave it open for any reasonable amount of time and it leaks memory every where and takes a year to shut down.

I'm sure their databases are nifty and everything but they're insanely expensive and I don't think they offer me enough over something like postgres to pay through the nose. I understand why corporations love it because they're all about having support even if it's not the best solution.

Re:Unsurprising

[Dwonis]

No, their databases *are* awful. You can't even store an empty string in a VARCHAR column.

Re:Unsurprising

[mcvos]

I'm not an expert on this, but apparently Java is quite a lot faster at allocating memory. And it can take advantage of runtime optimalization, which can make many common operations a bit faster in some situations.

Copying data around is probably going to be slower however, simply because of the memory overhead.

Re:Unsurprising

[magus_melchior]

Are you kidding? I'm sure you could fill them with individual departments.

Re:Unsurprising

[ToasterMonkey]

It might be a good idea at this point to start looking at other languages.

and

All you'd need is a portable virtual machine for it and you've a rival

really don't belong together.

Re:Unsurprising

[ToasterMonkey]

What is your list of cons?

Re:Unsurprising

[CynicTheHedgehog]

It wasn't until very recently (first with JSF, and then with Glassfish) that Sun got into the business of providing any kind of reasonable implementations. Apache has been the defacto RI producer for as long as I can remember:

Servlet/JSP: Tomcat
JSF: MyFaces
JAX-RPX/JAXWS: Axis
JAXB: JaxMe
EJB: Geronimo/OpenEJB
JPA: OpenJPA
JCR: JackRabbit
Java Logging: commons-logging, Log4j, SLF4j

(Those are just the ones I know off the top of my head). Apache was also a leader in developing most of these technologies with projects like Struts, Turbine, Velocity, Log4j/commons-logging, etc.

The JCP and open-source RI combination benefited Sun immensely. First, most developers had access to most tools and platforms for free, so there was no barrier to entry. The result? One of the largest developer bases for any platform. Second, developers had access to the RI source code, which means they can trace through the framework code, find/fix bugs, and offer their own improvements. The result? One of the largest support bases for any platform (BugZilla, Jira, forums, blogs, etc.). Finally, both of these offer tremendous incentives for commercial third parties to create their own offerings, which has lead to incredibly diversity: RedHad/JBoss, IBM/WebLogic, BEA/WebSphere, Oracle/OracleAS, Apache/Geronimo, Sun/Glassfish, and tons other.

And all of this creates a positive feedback loop that builds on itself.

So where did Sun benefit financially?

1. They receive licensing fees from the commercial third parties and in handheld devices (JavaME)
2. Certification fees
3. Hardware/software/support in their integrated platform offerings (admittedly this has suffered due to the popularity of commodity OSs like Linux)
4. (Indirectly) Not having to pay development staff to produce and support implementations

Oracle is attempting to dam the river so they can sell water to people downstream. If they are not careful, people are just going to find another water source.

I don't see this as the "end of Java", but I do see it as the end of the cohesive, constructive community built around Java. I expect to see fewer JSRs being ratified, more independent third-party libraries being used, and a lot of fragmentation in the community like we saw with web UI frameworks before JSF was introduced. I think it will be up to the container providers to try to hold it together and provide a consistent experience for developers.

Re:Unsurprising

[t2t10]

Good CS graduates have options, you know.

Re:Unsurprising

[t2t10]

Obviously, Java isn't free, as Oracle's lawsuit shows. As for hiring, yeah, you get your pick of 100 people, 97 of which are no good. In the end, you're no better off than if you pick some more obscure language.

The only people for who "I get tons of applicants" really works is very large companies, who have the resources to sift through all of them and the management structure to manage dozens of mediocre programmers to squeeze something out of them.

Re:Unsurprising

[Bigjeff5]

You are seriously underestimating what gets noticed and how far it goes up the chain. If they send 50 recruitment letters to a school and 15 of them come back saying "Not after you screwed Sun in the poop-hole, no way", they will definitely notice, and it will more than likely get pretty close to the CEO's desk.

That said, it won't change anything, and thinking it will is pretty naive of the GP. They've got money to offer, and 9 times out of 10 money wins out over slight moral quibbles.

Re:Unsurprising

[rve]

If one or more of these can be embedded into multiple browsers (IE, Firefox, Chrome, Safari and Opera would be the obvious set and cover almost the entire browser market), Java would face some serious competition - at least at the browser end. Java applications and servlets would depend on whether the Java ABI was covered by the patents. If the ABI (in and of itself) is not IP-protected, then it would be possible to write virtual machines that run entirely differently than "native Java" VMs but which support Java objects. Bring GCJ up to Java 7 and have a backend to GCC that supports a portable virtual machine. You then have something that will handle existing Java bytecode and will allow a gradual weaking off of Java to any language GCC supports.

(Since IBM -is- permitted to contribute to GCC, this is another direction IBM might be looking into. Especially if they can get a Java bytecode frontend working for GCC. Java applications natively compiled to IBM's processors would be very appealing, especially if it didn't break any standards in the process.)

For the first part, Java doesn't have a whole lot to do with browsers these days. I don't think it would cause major problems if applets disappeared completely. However, seeing as cobol, pl/1, rpg etc are still widely used in the business world, I don't think it is realistic to dream about Java being replaced in our lifetime. Sure, it will not be #1 for ever, but corporations around the world will keep depending on it for decades to come.

I don't think Java losing its #1 spot is much of an issue from Oracle's point of view. Even if a new policy causes half the Java customers switch to something else, they could probably still count on a massive increase in revenue if that means that the other half starts paying for it.

Re:Unsurprising

[janwedekind]

It's easy to look established in this economy.

Re:Oracle is Evil, C# Java

[h4rr4r]

C# is tied to a single OS. That makes it a nonstarter right there.

Mono is not a portable version, it is like its namesake a disease. Meant to poison the well that is Free Software.

Time for...

[markov_chain]

Geronimooooooo!

Re:Time for...

[dkleinsc]

They just need to call upon their superhero - Apache Chief Inyuk-chuk!

Re:Time for...

[mcgrew]

I'm picturing a skydiving cow...

Re:Oracle is Evil, C# Java

[SirGeek]

Keep in mind that Oracle Java is the "reference" copy of Java. Just because its the one most people use is not the point. For many moons, there was a couple of OTHER java implementations (Too bad Oracle now owns the BEA implementation of java too). There is still one that FreeBSD has (that was actually "blessed" by Sun).

" Apache Declares War On Oracle Over Java"

[RevWaldo]

Injuns, soothsayers, and volcanoes? Sounds like one hella cool game! When'll the demo be available?

.

Re:" Apache Declares War On Oracle Over Java"

[Arancaytar]

You left out the bit where the soothsayer ate a sun a while ago. Spectacular.

The Oracle at Delphi, Indigenous Tribes, Coffee

[Oxford_Comma_Lover]

> Apache Declares War On Oracle Over Java

Sometimes it seems like the world hasn't changed much in the last two thousand years.

Re:The Oracle at Delphi, Indigenous Tribes, Coffee

[Yvan256]

Except the arrows are now much bigger and go "boom!".

Re:The Oracle at Delphi, Indigenous Tribes, Coffee

[mtxf]

All of this has happened before...

Re:The Oracle at Delphi, Indigenous Tribes, Coffee

[aliquis]

So prior art is a valid counter-claim?

Re:The Oracle at Delphi, Indigenous Tribes, Coffee

[Bigjeff5]

Nah, he's just saying all of this will happen again.

Re:Oracle is Evil, C# Java

[sockman]

C# is better than Java anyway

Citation needed. For instance, Java has much better enum support (they are proper objects) from my perspective.

Reminds me of some bad history

[alexschmidt]

IBM-Oracle Pact. I wonder how long this non-aggression pact will last? This sounds suspiciously like the Molotov-Ribbentrop Pact. If you look it up, it had a really bad ending.

Re:Reminds me of some bad history

[jd]

Oracle wants to reduce the competition and to "leverage" IBM's expertise. Once that expertise has fixed the issues with OpenJDK and Harmony has died, IBM becomes disposable to Oracle.

IBM is most unlikely to stop all work on Harmony, they're just not going to distribute it. Oracle's implementation of Java will suffer performance and reliability problems. IBM already has its own compiler (Jikes) and IBM already has a Java distribution. Once IBM has the certification toolkit, it can internally continue to develop Harmony and upgrade Jikes to v7 Java. Remember, this is just a repeat of IBM's experience with Microsoft regarding OS/2 - only Oracle hasn't the muscle of Microsoft. Once IBM is satisfied, they dump Oracle, release their Java as standard on all IBM hardware and, because they have better ties with Linux than Oracle, on many Linux distros, and they'll likely be able to convince the courts that they don't infringe on any patents because they are officially licensed to be able to use whatever the technology is.

Again, though, IBM won't want too much competition in the Open Source community. They can't rob Oracle of power over Java if they aren't the de-facto controllers of Java. For now, they'll be best of enemies. Going back to the OS/2 fiasco, they learned the hard way that in such partnerships the first one to dump the other will be the winner. The partner left in the dirt WILL be trampled over, no matter how much better their product might be technically. And IBM will want to be the winner in this. Mind you, so will Oracle. Oracle will also be familiar with this process and will want to pull a Microsoft, killing IBM's Java work, forcing IBM to either sacrifice all they've spent or to sell it to Oracle at bargain-basement prices.

Re:Reminds me of some bad history

[mirix]

Are there really people that wouldn't know (at least roughly) what the pact was about without looking it up?

faith_in_humanity--;

Re:Reminds me of some bad history

[JAlexoi]

Well... IBM doesn't need to convince any courts of any patents violations. Because IBM can ask each employee to take a granted patent application and beat all of Oracle's employees to death with them :-D And there will still be granted patent applications left over for standbyers to join in mass-murder...

Re:Reminds me of some bad history

[jd]

This is high-stakes poker, with the winner claiming the cross-platform system as the prize. Yielding is getting dealt-in to the game.

If they play right, they can end up dumping Oracle, leaving Oracle in the dirt.

Or maybe the stakes are higher. Oracle and IBM are foes in many markets, and many of those markets now leverage Java. Whichever one is left controlling Java is also left controlling everything else.

To not yield (be dealt in), IBM would rapidly lose ground on its servlet engine (it would have no advance knowledge of how the specs are changing and no ability to ensure the specs benefit what they want to do). It could lose ground in the database arena (controlling the JDBC standard is valuable). And so on.

But if IBM gain control, by building a better Java on the sly and ensuring all the key systems use it at just the right time, then Oracle is in that boat. They now become the ones who lose control of servlets, JDBC, etc. That would wreck many of their key products.

This is a cut-throat business and these are two experts at throat-cutting rivals.

Re:Reminds me of some bad history

[jd]

IBM has a virtual machine, mostly compiled from Sun's sources with some optimizations of their own. And, yes, I have facts to back my claims, and yes, I have used IBM's JVM (I use it in preference to the Sun/Oracle one when I can).

Re:Reminds me of some bad history

[ckaminski]

The IBM VM is the default when running Websphere.

Re:Reminds me of some bad history

[TheTurtlesMoves]

Oracle's implementation of Java will suffer performance and reliability problems...

Now your just making crap up. They can't put the OpenJDK back in the bottle, and the current version has pretty good performance.

Re:Reminds me of some bad history

[TheRaven64]

IBM has several VMs. They have an entire research group at TJ Watson writing them. They then tend to fold the improvements back into their main commercial one, but they could easily take one of the research ones to market if required.

That said, writing a JVM is pretty trivial. A simple one is only a couple of thousand lines of code (one with decent performance is quite a bit more). The hard part is the class library, which is huge and must be 100% compatible with the Sun/Oracle version to be safe from trademark and patent infringement.

Nokia went for Python

[accessbob]

Nokia went for Python on Maemo. Looks like they knew what they were doing.

Re:Nokia went for Python

[mirix]

How so? maemo is full blown linux, you can have whatever you want on it... no?

or they have some default scripts on it that happen to be python?

Re:Nokia went for Python

[Anonymous Coward]

It's full blown Linux, but most users won't download an app which requires a huge runtime which doesn't come with the default install.

Re:Nokia went for Python

[Tester]

They went for C/C++. There is nothing in Python on the base Maemo platform.

For Meego, all UIs and higher-level stuff is in Qt so it's using C++

Re:Nokia went for Python

[accessbob]

As far as I'm aware, Java is still not officially supported on the N800 range. What has been supported from the beginning in 2004 is Python (PyMaemo) .

Re:Nokia went for Python

[horli]

Nokia has developed a LGPL python binding to QT since the existing pyQT is not fully open. Why should they do that if they don't count on Python? http://www.pyside.org/

Re:Oracle is Evil, C# Java

[h4rr4r]

Which part do you disagree with?

Re:Oracle is Evil, C# Java

[iONiUM]

That's a little dramatic, isn't it? Mono is the open source implementation of .NET, which is a very solid framework I might add, though clearly MS did wield it to further Windows (I don't deny that). Mono is released under GPL, LGPL, and MIT licenses http://en.wikipedia.org/wiki/Mono_(software).

I mean, facts are facts, so why do you have to be so dramatic about it? Or I mean.. did it.. where did Mono touch you? You can tell me.

Re:The Oracle at Delphi, Indigenous Tribe, Islands

[Anonymous Coward]

FTFY.

The Apache Indian in North America would go to war with the Athenian Oracle at Delphi over the island of Java in the South Pacific.

Re:Oracle is Evil, C# Java

[farnsworth]

Too bad Oracle now owns the BEA implementation of java too

BEA never wrote a JVM. They bought JRocket shortly before being acquired by Oracle.

Java is the new COBOL

[vlm]

Java is the new COBOL.

During the declining years of cobol, I/we watched the participants fighting to increase their portion of the pie, regardless of how much it shrunk the pie.

Re:Java is the new COBOL

[cacba]

Im confused, was cobol the only way to develop on millions of computer (aka smartphones)?

Re:Java is the new COBOL

[Rich0]

No - there weren't millions of computers to develop on back then.

However, cobol was the only way to develop anything that mattered on any computers that mattered. I wouldn't be surprised if the NYSE is still running on cobol and cics...

Re:Java is the new COBOL

[hsoftdev17]

If Java is the new COBOL, I highly recommend not telling the millions of Android developers out there, or Google for that matter. I am inclined to agree that the language formerly known as "Java" (Sun's version) may be on its way out. However, the existence of alternate compilers, alternate VMs, and extensions to the language not officially sanctioned by Sun (or Oracle) seem to indicate that Java isn't going anywhere anytime soon.

Re:Java is the new COBOL

[decipher_saint]

Let me put it this way, there are still tons of COBOL apps out in the wild. The last project I was on used a DB2 backend with a ton of COBOL stored procs. Imagine my surprise at having to learn enough COBOL to be dangerous in order to facilitate change to an application with an ASP.NET front end.

Re:Java is the new COBOL

[idontgno]

At one time, COBOL was the only way to develop on tens of thousands of computers. Very expensive computers with very expensive maintenance and licensing contracts. There was a lot of money in this, measured in hundreds of thousands of dollars per year per site. That's probably an order of magnitude or two lower than the money at stake in the mobile software universe, but it's also probably a larger percentage of the overall market at the time.

There is a common but entirely mistaken belief that the great issues and controversies of this time are unique, unprecedented, and never-before seen. But license and market-control conflict is ancient in this industry. Almost every hassle you may see today has been seen by some earlier generation of dinosaurs.

Re:Java is the new COBOL

[mark-t]

Although there is a noticeable similarity in the two languages in the sense of some of the politics involved, COBOL's biggest failure was ultimately tied to more technical limitations on the part of the language itself rather than the politics that governed it (and although those politics may have been responsible for COBOL's failure to evolve with the times, this does not appear to be happening with Java, at least not to the point that there are any foreseeable technical limitations).

Re:Java is the new COBOL

[leenks]

There are COBOL implementations that run on JEE servers too - double whammy if you are a COBOL *and* Java expert!

Re:Java is the new COBOL

[hey!]

Well, that's really a flawed analogy, because COBOL is a language and Java is really about *frameworks*. Of all those Java programmers out there, how many are working on projects that don't rely on some open source framework or library? That don't use Spring or Hibernate or Struts? That's not even counting Apache's Commons or Xerces.

When the fight comes, it's going to be completely different. It'll be Oracle trying to get a jiu-jutsu hold on a vast body of software using control of a small but strategic body of intellectual property.

Re:Java is the new COBOL

[SpaghettiPattern]

During the declining years of cobol

Cobol declining? Get real buddy. I for one see Cobol outliving Java.

And I'm not particularly overwhelmed by its beauty.

Re:Java is the new COBOL

[Pharmboy]

FORTRAN? While scientific computing is no longer the largest sector of computing, it is certainly something that traditionally "matters". And it was here before COBOL.

Re:Java is the new COBOL

[Raenex]

Java is the new COBOL.

People have been saying that for years, but what has replaced it? Is the pie really shrinking? What are new projects being written in?

Re:Java is the new COBOL

[JAlexoi]

When people say COBOL, they mostly mean CICS on mainframe.

Re:Java is the new COBOL

[gbjbaanb]

However, the existence of alternate compilers, alternate VMs, and extensions to the language not officially sanctioned by Sun (or Oracle)

sure, that makes sense. It worked fine for Microsoft when they wanted to add extensions to the language after all.

No, Java is on the way out, if there's one thing business doesn't like one bit, its uncertainty. Look at why Silverlight is now on the way out ('cos Muglia didn't support it nearly enough in his speech) and you'll see companies dropping it. Whilst a business is fine with paying a licence fee to use software, they cannot stand the thought of spending a lot of money on developing something that will be obsolete.

Java is now in the 'what the hell's going on' camp, and businesses are worried about it. You can guarantee they'll be looking at their future strategies right now, thinking 'what if' scenarios and (more importantly) looking at alternatives.

Re:Java is the new COBOL

[DiegoBravo]

> At one time, COBOL was the only way to develop on tens of thousands of computers. Very expensive computers with very expensive maintenance and licensing contracts. There was a lot of money in this, measured in hundreds of thousands of dollars per year per site.

Why you talk in past tense? How many mainframes are using Ruby for the business?

Re:Java is the new COBOL

[grouchomarxist]

I was surprised, but I recently met a COBOL programmer younger than me. Not only is COBOL still being used, new products are being created with it.

Re:Java is the new COBOL

[the_womble]

I would be very surprised. Stock exchanges pay for high performance, and their automated trading systems tend to be fairly new.

Re:Java is the new COBOL

[cstec]

If Java is the new COBOL, I highly recommend not telling the millions of Android developers out there

Not to worry, they don't listen to anything as it is!

Re:Java is the new COBOL

[Luke has no name]

Most banks in the United States are running some combination of Cobol and Java for the bulk of their back end/front end, respectively.

Re:Java is the new COBOL

[wzzzzrd]

Be surprised then ;) I worked on contract for IBM in the financial sector, and the very core of every bank, insurance corp or stock exchange I came across is some sort of transaction system (cics, ims) with a huge code base of COBOL packages. These systems run on shiny new big irons and the software is everything but ancient. Nobody risks to replace the core applications, they are the most stable systems I have ever seen and fast as hell. This is the sole reason COBOL will and cannot die in the foreseeable future.

Everything else runs on top of systems, they are the backbone of the financial world.

Re:Java is the new COBOL

[PybusJ]

All of that subject to successfully avoiding Oracle's lawyers, of course.

You do know that Google are being sued by Oracle over Android's java-like Dalvik VM. If that doesn't play out well then non SunOracle Java is pretty dead.

Re:Java is the new COBOL

[steelfood]

the existence of alternate compilers, alternate VMs, and extensions to the language not officially sanctioned by Sun (or Oracle) seem to indicate that Java isn't going anywhere anytime soon.

That's what Oracle is suing everybody and their moms over. Very soon, those alternate compilers, alternate VM's, etc. will be buried under a very large, hot, steaming pile of legal refuse.

Re:Java is the new COBOL

[magus_melchior]

Very true, but sometimes all it takes is an injunction against Dalvik or Harmony to really kill off even the alternatives in the marketplace (which Oracle would love more than anything). Should that happen, they'll exclusively become the purview of enthusiasts, which neither Google nor any Android handset maker wants.

Re:Java is the new COBOL

[t2t10]

FORTRAN wasn't a viable alternative to COBOL in the business world for technical reasons.

Re:Oracle is Evil, C# Java

[Darkness404]

Mono should be looked at like WINE, useful to port programs to, useful to get some programs to run, but shouldn't be your language of choice if you want to get cross-platform apps.

It's a trap

[rsborg]

Mono is a trap, and solely exists at Microsoft's pleasure. Once MS decides the want to kill it, out go the patent infringement lawsuits and anyone using Mono is on shaky ground unless they donate to Microsoft's coffers.

The fact that it hasn't happened yet is no insurance. Copyright/left is one thing, patents are another and I don't trust Microsoft.

Re:It's a trap

[Americano]

It's released under LGPL, GPL, and MIT licenses... how would Microsoft ever "kill" it? They may decide to no longer officially support it, certainly, and they could stop contributing future changes to the open source implementation... but serious question:

Once it's been released under GPL, how exactly could they sue someone for using it, or forking it and continuing to work on a parallel implementation? That seems like it wouldn't stand up for a single moment in a court.

Re:It's a trap

[zombieChan51]

Talk about paranoia

Re:It's a trap

[Enderandrew]

The are a couple problems with this theory.

The EU demanded that Microsoft open several of their standards and protocols, or else. The EU can stop the sale of Microsoft products in the EU and levy more fines.

And Microsoft has made an open patent pledge.

http://www.microsoft.com/interop/principles/osspatentpledge.mspx

If they go back on that pledge and tell the EU they refuse to cooperate with their demands on interoperability, then the EU hammer drops again.

Microsoft isn't going to do that. It makes zero sense.

Re:It's a trap

[h4rr4r]

You sue them over patents. Look at what MS is doing to folks build android handsets.

Re:It's a trap

[h4rr4r]

if You make or use such software outside the scope of creating such software code, You do not benefit from this promise for such distribution or for these other activities.

Create all the software you like, but if anyone uses it they reserve the right to sue them.

Re:It's a trap

[Enderandrew]

Novell extended the patent protection to anyone who uses their software. And again, if Microsoft tried to sue a company merely for using Mono software, the EU could come down again.

And I don't care who you are, a half billion dollar fine hurts. And stopping all sales of your products in the EU hurts.

Re:It's a trap

[amiga3D]

It's not paranoia if they're really out to get you.

Re:It's a trap

[Americano]

I've got to be honest, I'm still not seeing how that would fly in court:

Ballmer: "Here developers, developers, developers, use this Mono thing, it's awesome! You can even use our code!"

Developers Developers Developers: "Gee, Mister, Thanks! This is great, and it's even open source!"

Ballmer: "BWAHAHAHA! SUCKERS SUCKERS SUCKERS!"

Judge: "Well, Microsoft released all this code, actively supported the open source implementation of this product, and then changed their minds. I'll have to aware billions of dollars in damages for infringement to Microsoft because..."

It's that "because..." I'm having trouble filling in here. MSFT is claiming infringement on something they had no hand in creating with Android; how could they turn around and sue somebody for using something which *they helped create,* and which *they supported,* and which *they contributed source code to*? If they closed-source future revisions, somebody was bored enough to fork the GPL-released version and attempt to keep it in sync, and that forked version somehow implemented a new feature for compatibility, I could then see *some* possibility they'd go after the competing implementation, but I think if they changed the licensing on Mono back to closed, it would die from neglect soon enough anyway.

I just have trouble seeing how MSFT could successfully sue for patent infringement when they released & blessed the source code in its original implementation, at least not without a convoluted chain of events whose likelihood would seem to strain credulity.

Re:It's a trap

[vadim_t]

You've not really read that page, have you? Let me quote some bits:

If You engage in the commercial distribution or importation of software derived from an open source project or if You make or use such software outside the scope of creating such software code, You do not benefit from this promise for such distribution or for these other activities.

Meaning, if you distribute, or are doing it for money, it doesn't apply to you.

This promise is not an assurance either (i) that any of the Microsoft-issued patent claims cover a Covered Implementation or are enforceable or (ii) that a Covered Implementation would not infringe on patents or other intellectual property rights of any third party. No other rights except those expressly stated in this promise shall be deemed granted, waived, or received by implication, exhaustion, estoppel, or otherwise.

This is to be expected of course, but MS has an easy way out: sell the patents and have somebody else enforce them.

Here's another opinion

Re:It's a trap

[beldraen]

The very point of a patent is to publicly give out your understanding of an implementation for protection for a period of time from anyone else being able to use that implementation.

Specifically, if a person writes code that uses a patent, Microsoft cannot claim ownership of the code (because it is copyrighted), but Microsoft can sue you to prevent you from using the code because they own the idea.

Re:It's a trap

[MCEscher]

I was at the Mono breakout session at PDC 2008. The speaker had nothing but good things to say about Microsoft and the support they have given the Mono team. It was even a Mono breakout session about developing games for the iPhone.

Re:It's a trap

[ADRA]

"Novell extended the patent protection to anyone who uses their software."

So does this propagate with the GPL, LGPL, MIT based licensing that Mono apparently has? If I included a single file of mono and developed a clean room implementation independently does this indemnification still exist? What's the point of releasing something with an open license if nobody receiving said code can benefit from said license?

Re:It's a trap

[harlows_monkeys]

If only Microsoft would make a legally binding promise not to sue the standardized parts of Mono for patents. If only they would release many of the other parts under an open source license with a strong patent grant, like the Apache 2 license. If only they would take actions that would set up a very strong estoppel defense against suits over the rest...

Oh wait, they did all this. Go troll elsewhere.

Re:It's a trap

[Americano]

As I just pointed out... that's exactly what they're doing by suing Android - an OS they had no hand in creating - for infringement on patents. Reading comprehension fail?

I find it difficult to see what grounds MSFT could sue a Mono user on, considering they blessed, implemented, and *released* a great deal of the source code that goes into Mono. It would be like suing your past self - "Well, one time I thought it would be cool to let other people have this stuff, so I released it to them all. But then I changed my mind, and now this court needs to take money from those people and give it to me."

It would seem that Microsoft's damages would be self-inflicted. And I'm not certain I see much legal basis for them rescinding their covenant(s) not to sue, as well. I can't imagine that "well we changed our mind" would be enough for a court to declare someone guilty of infringement, when MSFT has made a public statement that they wouldn't sue.

Re:It's a trap

[Americano]

I understand how patents *work*, I'm having trouble seeing how Microsoft could release & bless source code implementing their patent, and then turn around and sue people for using that implementation.

Especially with the covenants not to sue that they've released in at least several of these instances where they're dabbling with open source.

It seems like this "OH NOES THEY'LL SUE" attitude is more based on fear & mistrust than any legal grounds that they could win a court case on. Not that fear & mistrust are necessarily misplaced, I just don't see how they could argue in court that some company using Mono would be infringing, when it's a project *they* had a huge part in creating, releasing, and supporting.

Re:It's a trap

[aristotle-dude]

Microsoft did not release any code to the Mono guys. They reimplemented .NET on their own. Because of this, MSFT can claim patent infringement.

Re:It's a trap

[JAlexoi]

Did they release the code for Mono under GPLv2 or v3? If not, then they are not bound by GPLs patent clause. Even if they "blessed"...

Re:It's a trap

[Enderandrew]

That is the thing. The Mono team owns the copyright to the code and they released it GPLv3. Microsoft would have a hard time claiming patent infringement when Microsoft worked with the Mono team and gave them their blessing.

Everyone is so convinced this is a massive trap, and it is actually insanely implausible.

Re:It's a trap

[Enderandrew]

No, if you try to make a commercial spin of that open software, you don't get the protection. But people usually don't sell GPL software. They sell support for GPL software.

And again, it doesn't matter because the EU has threatened them, and proven they will enforce that threat with the two fines they already levied.

On top of that, the Mono team fully owns the copyright to the Mono code and released it as GPLv3.

If Microsoft wanted to assert patent rights, they would have had to do so immediately. By letting the Mono project ship with a GPLv3 license and not objecting, Microsoft basically gave them their blessing.

In fact, Microsoft explicitly gave them their blessing by working with the Mono team.

What jury is going to listen to that trial and side with Microsoft that Microsoft was somehow wronged by people using Mono software?

Re:It's a trap

[Homburg]

Actually, I believe mono includes Microsoft's implementation of the Dynamic Language Runtime, so mono does include some Microsoft code. I'm not sure how this affects the patent situation - the DLR is licensed under the Apache license, which does include a patent grant, but I don't know how much this would cover. The Apache license grants rights to use any patents "necessarily infringed" by the software; I'm not sure whether or not the fact that the DLR runs on top of the CLR means that it "necessarily infringes" any patents Microsoft might have on the CLR (in which case the patent grant in the license would apply to all of MS's patents on the CLR).

Re:It's a trap

[slack_justyb]

Novell extended the patent protection to anyone who uses their software.

That's right, anyone who uses Novell's Mono. That right doesn't extend to anyone else.

The EU isn't going to "hammer" Microsoft when Microsoft has "given" ample opportunity to each major distribution to join them. That would make an easy case for Microsoft to make the EU look like the boogyman.

"Oh I'm so sorry that your copy of Windows is now 300 euro more than the last version. Why not call the EU and have them explain why they're forcing us to pass the cost on to the consumer when all we did was try to protect our contracts?"

People are idiots, governments doublely so.

Re:It's a trap

[gbjbaanb]

No, Microsoft licenced the patents to Novell. Novell then grants anyone using Mono these patent licences so they won;t be sued. That's it. If you get Mono from anywhere else, (eg Debian) then you are not covered.

Microsoft can sue your ass and still get to claim the moral high ground of giving it away for free yet still trying to control it from chaotic implementations and forks and whatnot. We may agree that'd be a stupid argument, but they'd probably win a court case using it.

So, is it a trap? Well, nobody in their right mind ever though writing Java apps would get them sued or have to suddenly pay up, yet here we are today. Stick to the truly open languages and stay away from the company-controlled ones, its about time we told them where to go.

Re:It's a trap

[penix1]

Judge: "Well, Microsoft released all this code, actively supported the open source implementation of this product, and then changed their minds. I'll have to aware billions of dollars in damages for infringement to Microsoft because..."

It's that "because..." I'm having trouble filling in here.

You are having trouble because you are confusing copyright law with patent law. That is a common mistake caused by all those attorneys who refer to both as "intellectual Property". To break it down to you, copyright != patent. Patent rights can be asserted any time by the patent holder regardless of their involvement with others. A patent holder can grant or withdraw licenses to use the patented technology at their whim. Currently, Microsoft only has a promise not to sue those developers that they can withdraw at any time.

The "patent pledge" Microsoft made with the mono developers can evaporate in the wink of an eye and then the whole project, and anyone using it, are at risk of patent infringement. The only reason they made that pledge is because of the multiple suits both here in the US and in Europe regarding interoperability. Once the scrutiny dies down, they can revoke that promise. And don't think they won't if it means they can make easy profits on threatening patent suits (which they can).

Worse, if Microsoft were to sell that patent to some other troll out there, nothing is binding that troll to the promise Microsoft made.

In short, you need to look up "torpedo patent".

Re:It's a trap

[Enderandrew]

No. The EU has hammer Microsoft twice and demanded they open their standards and protocols or they would be hammered again. The EU also made it clear they would halt the sales of all Microsoft products in the EU.

Why do you think Microsoft is playing nice now?

And how will Microsoft win a case in court when they didn't enforce the patent initially the moment a distro shipped a Mono product? Or the fact that they let the Mono team claim completely copyright on all Mono code and then license it GPLv3?

Microsoft worked hand-in-hand with the developers and helped them in that goal. They will have a real hard time convincing a jury they are suddenly a victim just because Ubuntu ships Mono.

Please stop spreading FUD.

Re:It's a trap

[Enderandrew]

Novell didn't license Mono specifically. They licensed the entire broad patent portfolio of Microsoft.

Given that Microsoft allowed the Mono team to license the Mono code under GPLv3 (given the patent clauses contained therein) and didn't enforce their patents when Ubuntu shipped Mono packages, Microsoft really won't have a case in court.

Stop spreading FUD.

Re:It's a trap

[FilthCatcher]

I agree with you on this point - using Mono you're pretty safe from being sued by Microsoft but there's one big issue with this whole patent issue that concerns me.

Currently there seems to be a bit of a patent arms race between a few large compaines. Most notably Microsoft and Oracle. Both these companies have a set of patents relating to VMs etc that seem to be fundamental to how these platfoms work.
The sheer number and breadth of these patents makes is look unlikely that there is nothing in Microsoft's offerings that voilates an Oracle patent and vice versa so we've got a cold-war style Mutually-Assured-Destruction stand-off in place.

The possible problem facing smaller implementations of either Java or .Net is that even if Mono gets an agreement from Microsoft mot to sue they are still vulnerable to being sued from elsewhere and they don't have their own stockpile of patents to act as a deterrant.

Re:It's a trap

[Raenex]

I just don't see how they could argue in court that some company using Mono would be infringing, when it's a project *they* had a huge part in creating, releasing, and supporting.

What are you talking about? Mono was written independently of Microsoft.

Re:It's a trap

[Americano]

The principle of estoppel would seem to apply here however:

Microsoft has promised not to sue, and stated so publicly, in writing. The Mono developers (and users) have proceeded under the assumption that commitment was made in good faith; Even if Microsoft reverses their decision, they cannot then sue for infringement of the *patents they already agreed not to sue* over. Estoppel would kick in, protecting the devs & users.

If the terms of the licensing arrangement change (at MSFT's decision, or because the project was spun off and sold to a patent troll), that might prevent me from *continuing to develop* the software and prevent me from using new releases because those new versions would not be covered by the patent covenant, but they'd have no legal basis for claiming damages on my 'infringement' on a patent which MSFT had publicly declared they would not sue over.

Such a change to licensing terms would likely kill Mono, and it would severely disrupt my business if I had strategic plans that included relying on Mono for the foreseeable future - i'm not arguing that reliance on Mono is a good thing, but I don't see how it approaches the level of "poisoning the well" of open source that the original poster suggested.

Re:It's a trap

[Americano]

Yes, and it implements the ECMA standards 334 and 335, both more or less written & submitted by Microsoft, defining the C# language specification and the Common Language Infrastructure (CLI) specification.

Re:It's a trap

[vadim_t]

No, if you try to make a commercial spin of that open software, you don't get the protection. But people usually don't sell GPL software. They sell support for GPL software.

That's tricker than it seems. My TV has Linux based firmware in it. I paid for the TV, so I was sold all that came in the box. I think it's reasonable to conclude that the firmware was paid for as well. There's lots of hardware with Linux on it these days.

On top of that, the Mono team fully owns the copyright to the Mono code and released it as GPLv3.

Sure, you own what you write, but that doesn't change patent issues.

If Microsoft wanted to assert patent rights, they would have had to do so immediately. By letting the Mono project ship with a GPLv3 license and not objecting, Microsoft basically gave them their blessing.

I see what you're getting at, but let me remind you of the GIF issue some time back. That it had been used for a long time didn't stop Unisys from asking and getting license money from some people.

In fact, Microsoft explicitly gave them their blessing by working with the Mono team.

That's for the court to decide. It's not a "You blessed them, you lose the lawsuit before it starts" deal in my understanding. It's a conclusion that may be reached after months or years of fighting in court.

What jury is going to listen to that trial and side with Microsoft that Microsoft was somehow wronged by people using Mono software?

What jury is going to listen to the trial? The jury that's selected for that trial, of course.

I agree that the way you say is how it SHOULD work if it goes to court, but reality is that if it goes to court the outcome is not 100% certain. Weirdness is possible, as well as out of court settlements. It may well drag on for years, while many people will be in a very uncomfortable situation meanwhile.

Re:It's a trap

[rsborg]

Novell extended the patent protection to anyone who uses their software. And again, if Microsoft tried to sue a company merely for using Mono software, the EU could come down again.

Fine. Great. Microsoft gets punished... well after my company and my career have been harmed, potentially fatally. The real problem is the corrupt patent office(s) and legal environment that allows companies like Apple, Blizzard and Microsoft to easily litigate what they want patent-wise.

Re:It's a trap

[Raenex]

First off, you've made several statements in this thread along the lines that Mono was a Microsoft project with Microsoft code. Obviously you were wrong. Second, Mono implements much more than the ECMA standards. They aim to implement all of .NET, including other Microsoft traps like Silverlight.

Re:It's a trap

[modmans2ndcoming]

sorry tin man. MS has no control over it. They have ruled out being able to sue it since they declared that mono can use MS patents dealing with .net and the CLI in perpetuity.

Idiots have been claiming that MS will drop the ax on Mono any day and it has never and will never happen.

Re:It's a trap

[walshy007]

Catch is microsoft have already signed an agreement not to sue open source implementations of c# and .net stuff

Re:It's a trap

[The End Of Days]

That is a common mistake caused by all those attorneys who refer to both as "Intellectual Property"

Picture if you will a person creating an HTML page. Picture that they don't really know much about computers, but consider themselves to be "programming" in some way. Is this Tim Berners-Lee's fault for specifying HTML to look, to the lay person, like a programming language?

I'd say any confusion about legal issues is more likely caused by the fact that being a nerd doesn't imply being an attorney, but it does imply acting like you know everything.

Re:It's a trap

[SplashMyBandit]

I think the term you want is, "submarine patent".

Re:It's a trap

[Requiem18th]

How to break your legally binding promises -- the legal way:

1. Never outright license.
2. Only promise not to sue...
2.1. Over patents you control.
3. Sell patents to patent troll and/or puppet company.
3.1. Voilà you no longer control the patents
4. Set us up the patent bomb.
5. Profit

There not even a fucking mystery "????" here.

Re:It's a trap

[Requiem18th]

Estoppal can be worked around, easily, sell patents to troll/puppet, have puppet do the suing. Easy as cake, there are not provisions against this. Or well there are, it's called granting licenses, which MS hasn't done about c#/.net

Re:It's a trap

[shutdown -p now]

The possible problem facing smaller implementations of either Java or .Net is that even if Mono gets an agreement from Microsoft mot to sue they are still vulnerable to being sued from elsewhere and they don't have their own stockpile of patents to act as a deterrant.

This is true, but note that it applies to any VM with JIT compilation - including such projects for e.g. Python and Ruby.

Re:It's a trap

[L4t3r4lu5]

The "patent pledge" Microsoft made with the mono developers can evaporate in the wink of an eye and then the whole project, and anyone using it, are at risk of patent infringement.

Isn't that commonly known as the "bait and switch"? Sucker guys in with promises of riches or benefit of some kind, change the rules, profit? Sounds... Hazardous.

Re:It's a trap

[CarbonShell]

+1

Not only do we not trust M$, we expect them to pull this kind of crap.
It is like running into the base during a TeamFortress game. You KNOW there is a trap, you just do not know where or when it will close.

Re:It's a trap

[CarbonShell]

So if M$ would sell the patents of C# to, dunno 'Microhard Patent Protection Company' and then THEY could sue for patent infringements and M$ can wash their hands in innocence.

Would anyone else doubt M$ might do this if they wanted to?

Re:It's a trap

[TheRaven64]

No they didn't. They agreed not to sue implementations of the ECMA specifications related to C# and .NET. These are a subset of the stuff that Microsoft created and Mono reimplemented.

Re:It's a trap

[thoromyr]

has microsoft paid a single penny of their EU fines to date? No? Have they complied with all EU requirements? No?

I don't think they give a flying fart about what the EU says. The EU threatened an injunction against sales, but it never happened. An injunction was attempted in the US (for MS illegally incorporating another company's IP into MS Office) and was shot down by a judge because it would be hard on Microsoft and its resellers. Really? That's kind of implicit with an injunction. But apparently Microsoft is "too big to fail" and so is above the law. If you really think that only applies in the US, note the complete lack of any real action by the EU against Microsoft. No seizing of assets, nothing. Complete failure to comply with EU court orders with no repercussions.

Microsoft does the bare minimum to appease those who have a modicum of power over them. They bought ISO approval for their closed office document "standard" because it helped confuse the issue over open document standards.

The "patent pledge" you link to is meaningless. It only covers a developer who plays by Microsoft's rules (it has to be a "covered implementation" for starters). It does not apply to commercial use of open source code (RedHat, for example, is a commercial distributor of a lot of open source software). It explicitly excludes anyone who does something in addition to coding for open source "covered implementation".

The "pledge" says "do what we want, how we want it, make no money (directly or indirectly), use nothing than our "open source" and we promise (corss our fingers) that we won't sue you.

Re:It's a trap

[Americano]

I made several statements saying that they had a hand in creating it (they created the .NET platform, published the standards) and supporting it - which they have, through covenants not to sue, and, through collaboration with the mono project on implementation of silverlight, as well as publication of numerous open source components.

My understanding was that the code from Microsoft's DLR was included in mono, perhaps that's not the case - in which case, it's still quite clear that Microsoft has still supported & contributed to the Mono project, even if they haven't directly coded it. Part of the Silverlight agreement gave the Mono project access to the full Silverlight test suite & specification to ensure full compatibility with the MSFT version, in fact.

None of what you've said has made the case for how Microsoft could suddenly just change their mind and decide to sue Mono & its users out of existence.

Re:It's a trap

[Americano]

Really, so the puppet/troll company could sue people for 'infringing' on a patent before they owned it, and where the current owner specifically allowed it and encouraged it at the time?

If MSFT sells .NET to a puppet, that puppet cannot revise history and say "We own it now, so you were infringing all along while you worked with Microsoft."

They can withdraw their support, they can sue forks of the product after they withdraw their support, they can make life difficult for users & developers (and tarnish their own image in the process), but they cannot sue you for something you did with the then-owner's blessing 5 years before they became the owner of the product.

This doesn't work around Estoppel, this changes the agreement - Microsoft would still be unable to sue for infringement during the times & under the conditions which they specifically encouraged people to perform that infringing activity. They could ONLY sue for infringement *after* they withdrew their support.

Re:It's a trap

[Enderandrew]

Fact. Microsoft worked directly with the Mono team and gave them their blessing.

Fact. Microsoft made a patent pledge.

Fact. The EU said if Microsoft doesn't play nice on interoperability then they will fine them again and halt all sales of Microsoft products in the EU.

Fact. Several distros and companies have shipped and used Mono software without licensing any patents and nothing bad has happened to any of them.

Fact. Microsoft has sued exactly one company over patents, and that was related to a file system.

Fact. Microsoft has a horrible track record as the company of Embrace, Extend, Extinguish, but in recent years Microsoft has opened documentation on their protocols, and actually released code to the OSS community. Their Chief Software Architect wrote in internal memos that Microsoft needed to change and start embracing open standards and protocols more. In that time frame, IE has shifted significantly to a standards compliant browser, and MS Office voluntarily added a filter for ODF and PDF. Not everything they do is some insidious plan.

Crazy Conspiracy Theory: Anyone who uses Mono will get sued to oblivion, even though it hasn't happened to this point, and it flies in the face of all the evidence to the contrary!

Yes, you are spreading FUD. Again, stop it.

Re:It's a trap

[Enderandrew]

Your scenario depends on two things.

1 - Microsoft would make so much money suing you that it would be in their best interest to halt all sales of their products in the EU, and pay another half a billion dollar fine. Do you have half a billion dollars that Microsoft can sue you for?

2 - Microsoft would win their court case against you when they worked with the Mono team, and allowed them to license the Mono code GPLv3.

http://www.fsfe.org/projects/gplv3/patents-and-gplv3.en.html

Re:It's a trap

[Enderandrew]

http://news.cnet.com/2100-1014_3-5255715.html

Microsoft did pay the EU fines.

And MS Office sales were halted. The judge upheld the injunction, not stopped it.

http://techie-buzz.com/microsoft/injunction-on-sale-of-microsoft-office-2007-word-2007-in-the-us-post-jan-11-2010-upheld.html

Then Microsoft swiftly resolved their patent case to resume sales because they are terrified of losing one of their two biggest cash cows. They can not afford to have an injunction against sales.

Please stop lying and spreading FUD.

Re:It's a trap

[Raenex]

You said: "It's released under LGPL, GPL, and MIT licenses... how would Microsoft ever "kill" it?" They may decide to no longer officially support it, certainly, and they could stop contributing future changes to the open source implementation... but serious question:

Once it's been released under GPL, how exactly could they sue someone for using it, or forking it and continuing to work on a parallel implementation? That seems like it wouldn't stand up for a single moment in a court."

Clearly you thought Mono was released by Microsoft under open source licenses. That's just not the case. Now you're backtracking. The DLR is a late addition and only part of Mono.

The other issue is that the Community Promise only applies to the ECMA standards. All the other crap that comes bundled with Mono isn't covered. Novell has a special deal with Microsoft with regards to Silverlight, which means Mono isn't open source for other Linux distributions.

Finally, even if Microsoft applied the Community Promise to all of .NET and Silverlight, they'd still screw you over. If you want vendor lock-in, follow Microsoft's lead. They will pretend they're interested in being cross-platform while trying to gain popularity, and then yank the rug out from under you on all future versions.

Re:It's a trap

[Americano]

Clearly you thought Mono was released by Microsoft under open source licenses.

No, clearly, I thought "IT" (being Mono) was released under LGPL, GPL, and MIT licenses. Which, surprise surprise, it *is*. I never said "MICROSOFT" released it under those licenses.

The DLR is a late addition and only part of Mono.

So then, I was correct? The DLR, written by Microsoft, *is* in fact a part of Mono? Or is it not?

Your rambling is incoherent. You are trying to attack points I've never made, and I also never said that "using Mono was a good idea." I simply don't see how anybody could open themselves up to legal action by using it.

You could certainly face business and financial challenges if you invest heavily in using Mono and Microsoft decides to withdraw its support from the project and kill future development of Mono, but I fail to see how anybody could open themselves up to a patent lawsuit over use of Mono as it exists today. Perhaps some future, forked-after-MSFT-withdraws-support version could be an issue, but right now, the 'it's a trap' brigade is doing nothing but spreading FUD.

There may be a host of reasons not to use Mono. "ZOMG MICROSORFT WILL SUE U WIT DA PATENTZ!!11" is not one of them.

Re:It's a trap

[slack_justyb]

I don't think anyone is paying attention this deep in a thread.

Who are you trying to convince? I know it's not me as you can see that I've made up my FUD spreading mind.

My point is that the deal extends to Novell and Novell users alone. I honestly don't care how Ubuntu, Fedora, etc... choose to slice, interpret, put into their default install, or whatever else Mono and the agreement/promises that Microsoft has made with Novell. The wording extends to Novell and Novell alone, the rest is just random guesses about how some company will act to other randomly picked companies/projects.

At any rate, Microsoft could be happy go friendly with open source, maybe not. I'm not the CEO so I couldn't tell you and I'm doubtful you can say with any amount of certainty exactly how Microsoft will act with respect to rightfully asserting their patents, if they should choose to do so. Maybe Microsoft won't, who cares? It's not like anything of critical importance depends on Mono, or that anyone would put anything of critical importance without getting some mass rebuttal from the FUDsters like myself, apparently. I really don't see Mono as any major issue at the current moment, nor do I see it as any major contention point any time in the near future, but that's just my humble opinion. I honestly think the whole Mono thing is just so that Microsoft can put a "cross-platform" label on their ads. I want to reiterate that, I believe Mono to be nothing more than just an ad helper for Microsoft, nothing more sinister or benign than that.

What I'm getting at is what will the EU do when Microsoft does something that they are legally entitled to do? Stop them from doing something legal? Punish them for doing something that the same body that will punish them, says that they are there to protect? That applies to more than just the limited scope that you are so intent on focusing on, namely Mono. It's a sticky situation for the EU and I don't believe "Hammer" is the correct wording, maybe more like "discuss very critically." For example how compliant is Office and the ISO OOXML standard? What's the EU doing about it? I'm sure there are some people who are taking care of that, and I'd like to believe that when the next version of Office comes out it will be fully compliant with the ISO OOXML standard so that we can have interoperability with that client.

Also, I have to really call out your ODF "Fact" because Microsoft's implementation of ODF doesn't work with other ODF implementations. Go ahead give it a whirl. FACT: Microsoft added voluntarily to Office an ODF filter that produces ODF files that are not compatible with other ODF implementations. Nothing insidious about that, just a bad implementation that's all. I'm sure someone else of a higher FUD-factor than I would assert that, "M$ is trying to fragment ODF!! OMG!!! blah blah blah."

There again, I'm not trying to spread any FUD here, just saying that the EU puts themselves into positions that are going to be difficult, maybe not that bad; maybe just uncomfortable, to explain to companies that wish to seek patent protection from that governing body.

The EU can do whatever they please, including whatever "Hammer" strategy that you are purposing. However, that's going to send a muddled message to other companies that want to assert/obtain patent protection from that body.

That's all I'm saying, nothing more, nothing less. It could be Oracle, IBM, Apple, or whoever, but you slap a person down for doing something they are legally entitled to do, and companies are going to start questioning how valid said protections are and to what degree they are allowed to exercise those rights. That, a good business environment, does not make.

Re:It's a trap

[Raenex]

No, clearly, I thought "IT" (being Mono) was released under LGPL, GPL, and MIT licenses. Which, surprise surprise, it *is*. I never said "MICROSOFT" released it under those licenses.

You asked: "Once it's been released under GPL, how exactly could they sue someone for using it, or forking it and continuing to work on a parallel implementation?". That question would make sense if it actually was Microsoft who released the code.

Oh, and there's more quotes from you. You're definitely backtracking:

"If they closed-source future revisions, somebody was bored enough to fork the GPL-released version and attempt to keep it in sync, and that forked version somehow implemented a new feature for compatibility, I could then see *some* possibility they'd go after the competing implementation, but I think if they changed the licensing on Mono back to closed, it would die from neglect soon enough anyway."

How could Microsoft change the licensing on the Mono project when they don't own it, and never did?

So then, I was correct? The DLR, written by Microsoft, *is* in fact a part of Mono? Or is it not?

We're talking about the whole Mono project here, not one late addition to the game. I wish you would stop pretending like you knew the situation. You clearly thought Microsoft was the owner of Mono.

I fail to see how anybody could open themselves up to a patent lawsuit over use of Mono as it exists today.

Easy, by using a technology included in Mono that isn't part of any patent grant. The Community Promise is for specific technologies, not a general grant for Mono.

Re:It's a trap

[Americano]

You clearly thought Microsoft was the owner of Mono.

Yes, of course, you're right. Despite never saying that, you're contriving excerpts of what I've said to mean that I said Microsoft owns it.

You are clearly the person with more time to spend on this nonsensical and pointless discussion than I am. Congratulations.

Re:It's a trap

[Raenex]

Not answering how Microsoft can change the license on a project it doesn't own. Yes, it is pointless discussion, since you can't admit your mistake. Congratulations.

Re:It's a trap

[Americano]

Okay, last try: Please quote for me where I said "Microsoft" could "change the license" on Mono?

I said they could withdraw their support for it, certainly.

I said they played a huge part in creating, releasing, and supporting it - and they have, if they wrote a large part of the standards (relevant ECMA standards) and products (.NET) it attempts to implement, they contributed code, and they published covenants not to sue, as well as entering into agreements with Novell;

I never said, or even implied, that Microsoft "owns" the code, or the license. In fact what I asked was how Microsoft could ever "kill" the project given that it was open-sourced under permissive GPL, MIT, and LGPL licenses. They can't retroactively change the license on code that's already been released that way, they can only change the license on *new versions* of that code.

So what you suggest simply underscores my question, which was: how would Microsoft kill an open-sourced product like all the chicken littles around here are claiming they will as soon as they get around to it?

The patent threat seems remarkably flimsy, given the probability of various estoppel & laches counter-claims. If Microsoft can't retroactively change the license, and lets Mono go on 'infringing' on its patents for years, and then suddenly decides "ZOMG THEY'RE INFRINGING SUE SUE SUE," Microsoft would then need to show WHY they took so long to sue (provided the Mono project & Novell had a lawyer who has spent more than 20 minutes in law school), and it would have to be a pretty solid reason to avoid a dismissal, since Microsoft has, one more time, encouraged, supported, and even contributed code to this project that it is now claiming "infringed" on its patents and which it is now suddenly seeking damages from.

Re:It's a trap

[Raenex]

Please quote for me where I said "Microsoft" could "change the license" on Mono?

"If they closed-source future revisions, somebody was bored enough to fork the GPL-released version and attempt to keep it in sync"

Going from open source to closed source is a change in license, which you then say might induce a fork. Of course, since Microsoft does not own Mono, there's no need to fork the GPL version, as Microsoft cannot close what they don't own.

how would Microsoft kill an open-sourced product like all the chicken littles around here are claiming they will as soon as they get around to it?

This question has been answered many times. By suing users and distributors of Mono over patent rights not granted in Mono.

The patent threat seems remarkably flimsy, given the probability of various estoppel & laches counter-claims

Quite the opposite, given all the software patent nonsense we've seen in practice. Patent trolls have been successfully suing companies for many years now, often waiting around until a company gets successful or a technology widely adopted.

Microsoft has, one more time, encouraged, supported, and even contributed code to this project that it is now claiming "infringed" on its patents and which it is now suddenly seeking damages from.

Microsoft has encouraged specific parts, and entered exclusive agreements with Novell in others. That's not a blanket permission for all of Mono for other distributors and users. Also, .NET and Mono are always adding new features, so even if you think laches will save you for an early version of Mono, it may not for last year's hot new feature.

Re:It's a trap

[Americano]

Going from open source to closed source is a change in license, which you then say might induce a fork. Of course, since Microsoft does not own Mono, there's no need to fork the GPL version, as Microsoft cannot close what they don't own.

Except for the parts that they wrote, which presumably, they would still own the rights to license, unless they've assigned the rights to the Mono project, or the FSF, or someone else. If they still retain those rights, they could close the source for future versions, which would certainly affect the Mono project.

And none of that is more than a distraction from the point that Microsoft *certainly* cannot sue somebody for copyright infringement of GPL'ed code. Which leaves us with the patent claims.

Laches & other forms of estoppel would almost certainly come into effect, or at least drastically reduce the damages available to Microsoft, given the fact that they have been:
a) aware of;
b) supportive of;
c) contributors to;
d) collaborators with;
the Mono project. Your argument about patent trolls does NOT apply here - those are companies that sit around, waiting for somebody to make money off the ideas they claim to hold patents for, and then sue them.

If they went out, and *ACTUALLY HELPED* people implement their patents, and then tried to sue them, you would see a significantly different court scene play out. The principle of laches is even more applicable when the delay in enforcing a claim puts the defendant *in a worse* position.

Microsoft cannot truthfully claim they were unaware of the Mono project; a patent troll can;

Microsoft cannot truthfully claim they notified the Mono project it was infringing; they have not, and in fact they have encouraged the Mono project to continue development of new features; A patent troll can certainly do this;

Microsoft has also issued public statements & documents indicating it won't sue over patent infringement in "certain parts" of Mono. It is not blanket coverage, but again, it sets up conditions under which Mono was inclined to proceed as a project - once again, estoppel becomes an issue here. You cannot "entrap" people into infringing - encouraging them, claiming you won't sue, and then sitting back to bide your time while those people wrote more and more infringing code would all be supporting factors for any defendant's motion for dismissal.

Nobody has been able to cite any legal grounds which would allow Microsoft to sidestep the estoppel & laches defenses, other than "of course they could, if they decided to just ignore the law and buy the court." There are HUGE differences between "patent trolls" and Microsoft's behavior in the case of Mono. The parallels and precedent are simply not there to support your claims that Microsoft would do the same thing all the patent trolls do.

Re:It's a trap

[Requiem18th]

They could ONLY sue for infringement *after* they withdrew their support.

So? That's enough to give a huge punch to open source.

Although, as you say, they can't revise history and sue you for your past infringement, they can start charging royalties for the continued distribution of infringing software.

This doesn't just affect new forks but Mono as well since they never really had a license to use the patents in the first place.

To reiterate, if MS granted an irrevocable royalty-free license over these patents, Mono (but not later derivatives) would be protected forever.

But MS never gave Mono license to use it's patents, which they could have done easily, but rather "promised not to sue", technically Mono *is* infringing on MS rights, it's just that MS barred itself from enforcing them (for the purpose of implementing .NET, you can't use the patents for anything else like a non compatible derivate, for instance, a VM for a new language).

This is not just a gray area, it's a white room with the light switch on MS hands.

Re:It's a trap

[Raenex]

Rather than continuing to go in circles, I'll stop here.

Re:It's a trap

[modmans2ndcoming]

No... I mean Tin Man... he is so paranoid he has wrapped his entire body in tin.

Re:It's a trap

[Dwonis]

What happens if MS sells the patents to a third party, who decides to sue? Would promissory estoppel apply in that case?

Re:It's a trap

[Americano]

Probably not, unless the covenants, etc. Were part of the terms of sale.

But even if they weren't, the purchaser couldn't sue mono users for PAST activities conducted under microsoft's agreements. They could go after "new" versions of mono, and withdraw their support from future versions, but the impact of that would hardly be something that would taint or break "open source" in general, as the "poison the well" expression suggests.

Re:It's a trap

[Rakarra]

It's that "because..." I'm having trouble filling in here. MSFT is claiming infringement on something they had no hand in creating with Android; how could they turn around and sue somebody for using something which *they helped create,* and which *they supported,* and which *they contributed source code to*? If they closed-source future revisions, somebody was bored enough to fork the GPL-released version and attempt to keep it in sync, and that forked version somehow implemented a new feature for compatibility, I could then see *some* possibility they'd go after the competing implementation, but I think if they changed the licensing on Mono back to closed, it would die from neglect soon enough anyway.

You're treating Microsoft as if it was some single, monolithic, unchanging entity, but executives come and go, and their philosophies and attitudes hugely affect how a company treats the various communities that use or work with that companies products and standards. Maybe Ballmer has no intention to bring patent lawsuits, but what if he retired at the end of the end of the year and someone aggressively anti-source takes over and decides it would no longer be in Microsoft's interest to allow Mono to exist?

IBM used to be the Evil Empire before Microsoft, then for awhile it seemed like it would be the business champion of Open Source. Now everything is murky there. Apple used to champion open source, even advertising the OSX kernel as being open. Steve Jobs used to talk about it. Now Apple is leading the charge in one of the greatest removal of developer and user rights since the company's founding.

You can hope that a company will show restraint and wisdom, but your only true protection is to not allow a company to have that type of power over your projects in the first place.

Remember, developers at SCO used to contribute to Open Source and GPLed projects too. That didn't prevent new management from coming in, looking at the books, and doing everything in the company's power to generate income regardless of whether it meshed with prior philosophies or not. And they did it with Microsoft's direction and money, one of the many reasons why many around here would be a bit distrustful of MS's benevolence.

For a current example, look at what's happening with the whole Oracle/Java debacle.

Re:It's a trap

[t2t10]

Look at what MS is doing to folks build android handsets.

So you use Java, you still have both Microsoft and Oracle going after you. Obviously, using Java didn't help one bit.

Actually, if Android were based on Mono, there'd be less of a problem because Microsoft's legal commitments regarding Mono are much broader than Sun/Oracle's legal commitments regarding Java.

Re:It's a trap

[Bigjeff5]

Uh, you realize Mono and the patents it depends on are licensed under several open source licenses, right?

Your whole post is completely irrelevant.

Re:It's a trap

[t2t10]

Of course, there would have to be any actual patent infringement in the first place...

Re:It's a trap

[Requiem18th]

"Mono and the patents it depends on are licensed under several open source licenses"

If you mean "Mono is licensed under several open source licenses" than thats probably true. But I haven't been presented with "open source [sic] licenses" for .NET patents, all I've found, and all .NET shills have presented me with, is the "promise not to sue" with very limited scope and blobs of small prints that I neither want nor have to read.

Groklaw, the EFF and the FSF already read them for me and they say it's not safe and why. Let the professionals do their work.

Re:Oracle is Evil, C# Java

[Nadaka]

c# has some nice friendly language features. But the JVM still vastly outperforms the CLR runtime.

Re:Oracle is Evil, C# Java

[GreatBunzinni]

What the Java world desperately needs is a international standard defining both the language but the set of libraries and even the byte code that goes with it. As soon as the Java world starts following an international standard (or sets of standards) instead of being held hostage by a multinational corpotation then all these petty problems will go away.

Re:Oracle is Evil, C# Java

[h4rr4r]

The license is not the issue, the patents are the issue.

If it gains traction rest assured MS will come seeking rent like they trying to do with android now.

Re:Oracle is Evil, C# Java

[leenks]

No, it is an attempt at an open source implementation of .NET (which as you say, is a solid framework, despite all the COM crap) - which isn't much different that GCJ or Harmony - neither of which anyone that uses Java takes seriously as alternatives to parts of the Java stack (though they definitely respect the efforts and intents of the developers).

Mono is definitely NOT a .NET implementation, and nobody can use it as such in "business". And like GCJ or Harmony, no .NET developers really take it seriously for the same reasons, and the fact that most .NET developers are Windows people and couldn't care less about Linux.

However, mono though it can be used as "mono" in its own right if everyone understands the implications that brings. Many people prefer other offerings given that scenario however.

Re:Oracle is Evil, C# Java

[trelony]

Mono violates the same patents as Android and Harmony. Microsoft bought its licence from SUN, but it does not cover Mono. Unless Microsoft makes Mono its own project, it is no better than anything else. And I thought "rewriting Hudson in C is a stupid idea". Now it makes sense...

Re:Oracle is Evil, C# Java

[zombieChan51]

I think he disagreed with this

"it is like its namesake a disease. Meant to poison the well that is Free Software."

Re:Oracle is Evil, C# Java

[leenks]

No, Sun (now Oracle) defines Java and licences the trademarks and compatibility kit (ie the tests that you must pass in order to declare you are Java).

At the moment at least, nobody cares if you are a 100% complete Java implementation if you can't call yourself Java. You will not be adopted by most businesses, wont gather a large user base.

BEA don't have (didn't have?) an implementation of Java either...

And the BSD one is Sun's anyway.

Re:its on s60 as well

[accessbob]

It's a shame BlackBerry (aka RIM) haven't gone down the same route - they've tied themselves into a flavour of Java with a non-standard graphics API.

Re:Oracle is Evil, C# Java

[h4rr4r]

In what way?
Do you not think it exists solely to get MS patents into the free software ecosystem?

Do you think MS is just going to let it thrive ever?

Mono is like moonlight, it gets MS patents into free software land and lets them claim cross compatibility without any actual cross compatibility.

Re:Oracle is Evil, C# Java

[AndrewNeo]

I prefer having the potential to shoot myself in the foot using C# instead of always shooting my foot, my leg, and my neighbor's leg using Java.

Change this to an inflammatory title

[VGPowerlord]

If you wanted to make this headline more sensational, you could change it to "Apache says GPLv2 license not good enough." which is what OpenJDK7 is licensed under.

Yeah, Apache may be at war with Oracle now, but this has the potential for much more widespread damage. It also puts the Free Software Foundation in an... interesting position, as this technically is the first salvo from Apache in a license war between GPL and Apache License.

Re:Change this to an inflammatory title

[sproketboy]

Yup. Sad that on slashdot I had scroll through a hundred stupid tin-foil-hat comments to find the only one worth anything. And of course I have no mod points....

Re:Change this to an inflammatory title

[Homburg]

I don't think that's right; this isn't about GPL vs. the Apache license. The issue isn't the licensing of OpenJDK itself, but about the licensing of the Java Technology Compatibility Kit (the JCK), which is used to test if an implementation is compatible with a given version of the Java spec. The JCK isn't available under an open source license at all. If the JCK were under the GPL, or even if it were under a license that didn't permit you to modify it, but only permitted anyone to run it, then Apache could use it to test their Java implementation, which is what they want to do.

Re:Change this to an inflammatory title

[Anonymous Coward]

You don't know what you're talking about.

The problem is that to be a compatible Java implementation you must pass the TCK. To get a hold of the TCK you must agree that your Java implementation has a limited field of use, namely desktop computers. That means you have to add a clause to your licence that tells your users where they can use the software - no such clause exists in any open source licence I'm aware of.

Sure you can use the OpenJDK, you can even fork it, but therein lies the problem... you can't, because if you do and you want to claim it's a compatible implementation you have to pass the TCK. So you have to licence the TCK, then you have to add a field of use restriction to your licence, but that's incompatible with the GPL that the OpenJDK GPL requires you to licence under.

End result, you can have Oracle Java or 'Open'JDK

The ASF don't have a political axe to grind with the GPL, aren't firing a salvo in some imaginary war based on their view of free; It's about a contractual obligation Oracle has to release the TCK to the ASF. An obligation Sun had and failed to meet and that Oracle continues to fail to meet.

The ASF was re-elected to the JCP with 95% of the vote. No other elected member had anywhere near that. The members spoke with their vote and consequently the ASF leaving the JCP would be big news in a war with Oracle, nobody else. The ASF is outside core Java and the work of the JCP probably the biggest single contributor to the Java ecosystem. Their threat to leave the JCP would seriously damage it and Oracle's commitment to opensource's credibility.

Re:Change this to an inflammatory title

[jonwil]

Oracle (and Sun before them) said "if you build an implementation of Java that passes the JCK, you get a license to the Java patents". Then they said "we will license the JCK for testing on any implementation excpet for those that are designed for mobile devices"

Oracle of course wont give in to Apache on this.
Hell will freeze over before Larry will allow ANY implementation of anything that even vaguely resembles Java to run on anything vaguely reselmbling a mobile phone device unless the vendor shipping that implementation pays Oracle per-unit royalties for every device they ship.

Re:Change this to an inflammatory title

[cstec]

this has the potential for much more widespread damage. It also puts the Free Software Foundation in an... interesting position, as this technically is the first salvo from Apache in a license war between GPL and Apache License.

www.fuckgpl.com -- the shot was fired a long time ago. FSF - "Free as in Not" (tm)

Re:Change this to an inflammatory title

[struberg]

You have no idea.
The problem is that the Java spec says you only get all the IP and patent licenses granted if you pass the TCK in 6 months time from releasing a beta. But you have no way to access the TCK if Oracle doesn't give it to you! And despite Suns promise to do so, they refused to give it to us (nor to anyone else who doesn't sign a completely restrictive NDA and Field of Use restriction).

Fazit: even if the code is under GPLv2, you cannot simply fork OpenJDK currently! (At least not without getting your pants sued off you)

Ross and I recently talked with the chief of the FSFE at a conference and showed him the Java spec and Java TCK license wording and the problems with that. He promised us to talk with RMS about it because the might use the word GPL in the TCK-license in a misleading way. Btw, there is a good reason why there is now a GPLv3 which has an explicit IP and patent grant clause (as the ALv2 has). So there is no war between FSF and ASF but exactly the opposite.

LieGrue,
strub

Re:Change this to an inflammatory title

[mcvos]

What I'm wondering is how hard the line between mobile and non-mobile devices really is. A netbook doesn't count as mobile, does it? What about a netbook with a touch screen? What about a netbook with built-in UMTS? What about a netbook with both? And a keyboard that slides out from under the screen? Maybe make it a bit smaller?

We need more devices between iPads and netbooks, and use them to challenge Oracle's silly restriction.

Re:Change this to an inflammatory title

[mcvos]

Besides, shouldn't the standard JVM for linux work fine on Maemo?

Re:Change this to an inflammatory title

[TheRaven64]

The difference was the set of class libraries. They had Standard Edition and Mobile Edition, and the Mobile Edition stuff was expensive. It also came with an interpreter, rather than a JIT compiler, to reduce memory usage. Older ARM chips had extensions that let them run a subset of JVM bytecodes directly and trap to the VM for the ones that were too complex to implement in microcode, and the JVM for using these was also expensive.

Re:Change this to an inflammatory title

[oh_my_080980980]

Sun had no obligation.

The simple fact the open source community has it's knickers in a bind because Java is not completely, 100% free of restrictions.

TOUCH TITTIES.

The fact of the matter is Sun created Java and Sun watched over Java and protected Java. You remember Microsoft right???

Do you seriously think that Java would have survived if Sun opened up Java from the start?

Have you paid attention to what Microsoft has done to ODF and the way they pushed through their open document format through the EU process to get it to be a "standard."

Where's the advantage to being open again?

FYI businesses don't give a flying fuck if Java is open or not. They only care if their apps work. They only care if they can get the support.

Right now Oracle ensures that happens.

Re:Change this to an inflammatory title

[VGPowerlord]

Last I checked APLv2 was compatible with GPLv3, but *NOT* 2, due to the patent protection clauses.

Coincidentally I'm pretty sure the FSF wouldn't think badly of that complaint since the FSF itself recommends v3 over v2 for EXACTLY that reason. And given that Oracle is suing Google over those very patents, I'm pretty sure everyone would feel safer with oracle stating in writing that anyone using the code would not be subject to patent lawsuits. But that may just be me. YMMV and all that.

GPL compatibility is one-way. That is, you can take code from APLv2 and pull it into a GPLv3 project. You can't do it the opposite way around, because that would be a violation of the GPLv3 license, clause 5c.

Re:Change this to an inflammatory title

[ToasterMonkey]

The problem is that to be a compatible Java implementation you must pass the TCK. To get a hold of the TCK you must agree that your Java implementation has a limited field of use, namely desktop computers. That means you have to add a clause to your licence that tells your users where they can use the software - no such clause exists in any open source licence I'm aware of.

Why should open source software get industry certifications for free.. in any sense of free?

That is the real question here.

"Since August 2006, the ASF has been attempting to secure an
acceptable license from Sun for the test kit for Java SE. This
test kit, called the "Java Compatibility Kit" or "JCK", is needed
by the Apache Harmony project to demonstrate its compatibility
with the Java SE specification, as required by Sun's specification
license."

Read carefully, Apache needs what? Java certification, with no strings. Because, you know, open source just isn't open if you can't get it certified.
If UNIX certification requirements stated "you can't use the color purple on Mondays" nobody would bother certifying their Linux distributions as UNIX. OH WAIT, NOBODY DOES THAT ALREADY, BECAUSE IT'S OPEN SOURCE AND WE DON'T GIVE A SQUAT ABOUT CERTIFICATIONS. Unless you're Apache, then it's OK because Slashdot says so.

Re:Oracle is Evil, C# Java

[david_thornley]

The language devs at Microsoft few in a couple of nice touches (e.g. shorthand for getters/setters to reduce clutter in class definitions)....most of it is just a shortcut to bad software engineering.

I've got news for you.

Getters and setters are bad software engineering in general, so I'd have to consider properties to be a real fast shortcut to bad software engineering.

Re:Oracle is Evil, C# Java

[utahjazz]

GP: Mono is not portable
P: Yes it is! Here is a link to its license!

Being open source doesn't make something portable.

Re:Oracle is Evil, C# Java

[i.of.the.storm]

What about functional programming then? Java kind of blows if you want to use that programming paradigm.

No WaY!

[cez]

How could you say that! Why just today I got an email from them offering me $19,450 SAVINGS for Java Classes!

Get $24,450 worth of Java training for only $5,000!

Java Pass puts you on the fast track to become a certified expert Java programmer or developer.

Oracle University's Java Pass is a flexible education subscription program giving you access to Oracle's top ten Java courses at a fixed annual fee that stretches your training dollars. Learn at your own pace with these Live Virtual Classes while eliminating travel expenses, saving you time and money.

Live Virtual Classes from Oracle University

Live Virtual Classes are comparable to our traditional in-class training, but without the need for expensive travel. With the latest in collaborative technology, top instructors, cutting-edge curriculum, and hands-on labs, we offer an exciting combination of traditional content and interactive learning. Learn more.

What’s included in Java Pass

Java Pass gives you a full year of access to these top-10 Java Live Virtual Classes that prepare you for the latest entry-to expert-level certifications from Oracle University: Fundamentals of the Java Programming Language, Java SE 6 Java Programming Language, Java SE 6 Developing Applications with the Java SE 6 Platform Developing Applications for the Java EE 6 Platform Developing Web Applications Using JSF Technologies Web Component Development with Servlets and JSPs, Java EE 6 Building Database-Driven Applications with JPA Business Component Development with EJB Technology, Java EE 6 Developing Web Services Using Java Technology (COMING SOON) Developing Secure Java Web Services, Java EE 6 Java Pass also includes a discounted voucher for an Oracle University Java Certification exam to help you achieve your Java programmer or developer certification. This is an introductory offer that won’t last forever, so sign up now and save up to $19,450 on Java training! Terms and Conditions apply; contact an education sales representative for details.

To find out more about Java Pass, call 1.800.529.0165 (US) or 1.866.825.9790 (Canada) or click here.

Learn Technology from the Source

From initial deployment to fine tuning, from optimizing workflow to developing new applications, Oracle University offers you an unsurpassed selection of training options for your technology needs. For more than thirty years we have provided enterprises and individuals with training of the highest quality and value. Choose Oracle and learn from the teams that created the products.

Would a company who only had money in mind off such InSaNe SaViNgs!~!!

Re:Oracle is Evil, C# Java

[diskofish]

-1 Troll

I've had Mono running on a few different flavors of Linux, OS X, Windows, etc. It's portable enough. Portability is a misnomer anyway. It's not as if I can just take Java code I've written for Android and it will magically work on Windows.

Re:Oracle is Evil, C# Java

[diskofish]

You must be new here.

Re:Oracle is Evil, C# Java

[leenks]

Most Java programmers can't program Java either, so it doesn't surprise me that books on Csharp get it wrong either!

Re:Why be surprised?

[Yvan256]

Oracle: I'd ask you to sit down, but, you're not going to anyway. And don't worry about that language.
Apache: What language?
[Apache turns to look for a language, and as he does, he knocks Java, which shatters on the floor.]
Oracle: That language.
Apache: I'm sorry--
Oracle: I said don't worry about it. I'll get one of my intern to fix it.
Apache: How did you know?
Oracle: Ohh, what's really going to bake you're noodle later on is, would you still have broken it if I hadn't said anything?

Re:Oracle is Evil, C# Java

[VGPowerlord]

Getters and setters are bad software engineering in general, so I'd have to consider properties to be a real fast shortcut to bad software engineering.

Getters I could agree with, but setters make sense in that I can stop you from unilaterally changing my objects' variables to values that doesn't make sense/cause crashes/etc...

Re:Oracle is Evil, C# Java

[Panaflex]

Java has a huge tool base and lots of great development already done, more developers and a larger platform base. But, IMHO C# is a cleaner and more consistent language. Delegates are nice at times as well, something we're used in messaging and protocol libraries with great success.

I was about to say

[blair1q]

I was about to say "how the hell is Sun still in business?" for about the thousandth time.

Then I remembered...

Re:Oracle is Evil, C# Java

[BBTaeKwonDo]

Too bad Oracle now owns the BEA implementation of java too

BEA never wrote a JVM. They bought JRocket shortly before being acquired by Oracle.

I wouldn't consider six years to be be "shortly." Quoth the Wikipedia entry for JRockit: JRockit, a proprietary Java Virtual Machine (JVM) originally developed by Appeal Virtual Machines and acquired by BEA Systems in 2002, became part of Oracle Fusion Middleware in 2008.

Define Portable

[nurb432]

Mono runs on windows, mac, FreeBSD, linux and iOS. What more platforms would it need to be considered portable?

Not saying its perfect or there aren't other options, but at least be accurate.

Re:Define Portable

[aristotle-dude]

Mono runs on windows, mac, FreeBSD, linux and iOS. What more platforms would it need to be considered portable?

Not saying its perfect or there aren't other options, but at least be accurate.

Yes, command line stuff works fine but try using Monodevelop on OS X. Monodevelop is for all intents and purposes linux specific. It is a joke on any other OS because of the linux library dependency hell.

Re:Define Portable

[modmans2ndcoming]

http://monodevelop.com/Download

seems to show an OS X build.

Re:Define Portable

[aristotle-dude]

http://monodevelop.com/Download

seems to show an OS X build.

That is precisely the version I tried most recently. You cannot build any of the GUI template solutions other than the glade one and there is no GUI editor in the OS X version because of a bunch of GTK dependencies.

Re:Define Portable

[TheRaven64]

I'm not a fan of Mono, but you're placing the blame in the wrong place there. A lot of apps that use Mono use GTK#, which is a set of C# bindings to GTK, which is a library written in C (with an attempt at implementing C++ in macros). These will be unlikely to work on OS X, because the GTK port to OS X is useless. The version that uses X11 uses X11 (so doesn't work well) and the version that uses Quartz doesn't work at all.

It's akin to complaining that Java is not portable because an app uses Win32 via JNI and so doesn't run on Linux.

Re:Why be surprised?

[blair1q]

Bruce Almighty?

Dramatic fits the context of this article

[pizzach]

Most people never thought Java would become a hot potato to be careful with. No one thought that Oracle would be going after people over patents. Sun put Java under the GPL2.

Can you tell me how Mono is more safe being under the GPL/LGPL/MIT when it is using tech directly from a company that is in many ways a direct competitor and has outwardly stated it thinks of open source as "communism"? Microsoft does have patents on specific things used in Mono. Mono is also under the GPL2. Coincidence? I think not.

It's called a can of worms. It's just we have a lot of slashdotters who refuse to believe it now for whatever reason.

Re:Dramatic fits the context of this article

[Late Adopter]

It's not more safe, but after Oracle's actions it's no longer more dangerous. The patent issue used to be a very good reason to avoid Mono, but now for the feature space that Java and .NET occupy, there's no longer a good reason to prefer either of them!

For desktop applications you can tell people to start using Qt (does Nokia have any patents we need to be aware of?), but in the enterprise space is there really anywhere comparable to go?

Re:Dramatic fits the context of this article

[t2t10]

Most people never thought Java would become a hot potato to be careful with. No one thought that Oracle would be going after people over patents. Sun put Java under the GPL2.

Well, most people are obviously fools that know little about patents, licenses, or copyrights. What has happened was totally predictable from Sun's behavior and patents.

Can you tell me how Mono is more safe being under the GPL/LGPL/MIT when it is using tech directly from a company that is in many ways a direct competitor

I can tell you that if you look at the legal situation, Mono is and has been a lot safer than Java.

Re:Why be surprised?

[blair1q]

Oh. No. Sorry. Thought of my second guess the second I guessed.

But would I have if I hadn't made the first guess?

Re:The Oracle at Delphi, Indigenous Tribe, Islands

[H0p313ss]

The Apache Indian in North America would go to war with the Athenian Oracle at Delphi over the island of Java in the South Pacific.

Sounds like a game of FreeCiv

Re:Oracle is Evil, C# Java

[Dog-Cow]

The OP stated an opinion. A citation would be meaningless. And enums are not objects in Java, though they must be declared within an object. Which is stupid, as enums are types, just like objects are.

OpenSolaris Board commits seppuku redux

[khb]

Apache to Oracle: Do what we say or we'll resign!
Oracle to Apache: Sayanora

I don't know that they should stay, but if they want to have any influence working with Oracle, aligned along Oracle's self interest is the only way to have impact.

Declaring "war" and making threats is highly unlikely to cause any useful change in Oracle's direction.

Surely the OpenSolaris experience illustrates just how Oracle behaves w.r.t. threats.

Re:OpenSolaris Board commits seppuku redux

[OdinOdin_]

But Apache has (recently) been re-elected into the Java Community Process (JCP), along with Red Hat over the past weeks. The Oracle friendly candidate was rejected by the JCP. This issue was always due to play out in the JCP as actions to inhibit that process have been taken in the past.

So what may happen is the JCP process as we know it now might break down, where Oracle will not be allowed to act or make decisions due to democracy and getting out-voted by proposals from other JCP members. So what will Oracle do then, declare the JCP process void because no one will pass them the ball ? This only continues to segregate and divide Oracle from the rest of the pack.

If the JCP process breaks down, won't that be bad for Oracle stock price ? A few careless whispers from a dark corner when in that predicament, about the next big thing and how big-business will address compatibility with the language formally going by the name Java and suddenly the elected group have the backing of the public to start a revolution and dethrone the King. Its the stock price (or lack of it) that counts!

Re:OpenSolaris Board commits seppuku redux

[shutdown -p now]

Thing is, Java isn't just a language and a VM. It's also an ecosystem, and unlike that around .NET, it has historically being a much more open and diverse one. Case in point - the most popular Java build tool (Ant), Java ORM (Hibernate) and Java web framework (Spring MVC) are all third-party products. So Oracle has control over parts of that now, but by no means they control all of it (except in the "he who can destroy a thing, controls a thing" meaning of it).

Now, of all the players in this Java ecosystem, Apache has always been one of the major guys. Ant, Maven, Struts, Tapestry, Tomcat, Velocity... sure, it's all FOSS and others (like Oracle) can fork it, but do they have the resources to maintain all of it?

And if they let that ball drop, there goes a huge chunk of the ecosystem - and with it, a lot of what makes Java attractive today.

So, no. It's not nearly as one-sided as it seems. Of course, if Oracle just wants to kill Java as one of the most broadly used software platforms of all times, and make it into a .NET-like in the Oracle software stack with all tech coming from the house, then sure, they can do that. But I very much doubt they'd be able to monetize that efficiently. For those who don't mind that sort of thing, there is already .NET, and it has much more shiny in it than Java ever did.

Re:OpenSolaris Board commits seppuku redux

[gstein]

Apache tried that game for many years. Work *within* the JCP, and as a member of the EC. Apache *has* made many useful changes to the JCP and how it operates. But even with all that, it hasn't produced the JCK without an FOU clause.

At some point, enough is enough. The Apache brand name cannot continue to be associated with a sham.

Re:OpenSolaris Board commits seppuku redux

[snookiex]

OK, I know this movie. It's called "Resistance is futile".

Re:OpenSolaris Board commits seppuku redux

[ToasterMonkey]

Why is it so important to Apache that any Java implementation be certified with an open toolkit, for free?
I'm sorry, you have the specification, reference source, and free reign to modify and redistribute that. Why should certifying it as "Java" come without a price?

We can draw parallels to other projects very easily. Solaris is (mostly I think) open source, but the POSIX certification is not, and not free. Should OpenIndiana care?

Why does an open source foundation demand free industry certification? ??!

Apple can afford to buy Oracle

[Anonymous Coward]

...and put Java and MySQL back into open source - PERMANENTLY!

Re:Apple can afford to buy Oracle

[Dunbal]

Yes because the closed source, closed minded "you can't run that on my devices let us tell you the Apple way of thinking" Apple is all about Open Source.

Get real.

Re:Apple can afford to buy Oracle

[Builder]

Just out of curiosity, do you actually know anything about Apple, or do you just like to repeat what you've heard? Apple actually make significant contributions both to open source projects and international standards. Just because they lock down _some_ things, doesn't mean that this is their attitude to everything.

Re:Apple can afford to buy Oracle

[mcvos]

If you want help from big companies, Google and IBM are our only hope, really. And IBM already joined the dark side.

Re:Apple can afford to buy Oracle

[Vectormatic]

i was just wondering about that, how much cash would one need for a hostile take-over of oracle? and does google have that amount of cash?

i would have prefered Google buying sun right away, but if they happen to take down ellison on their way, that;s fine with me

Re:Apple can afford to buy Oracle

[mcvos]

Google tells me Oracle is worth $140 billion (twice what it was in 2006, so they've been growing well). Google will easily pay one or two billion to pay a company, but 140 is quite something else.

Google's market value is $150 billion, marginally more than Oracle.

Late 2008, Sun's market value had dropped to $3 billion, and Oracle bought it for $7.4 billion. That's something Google could easily afford.

Blame Sun, not Oracle

[harlows_monkeys]

The policy of lying to Apache about Java was started by Sun, not Oracle.

Re:Blame Sun, not Oracle

[TheTurtlesMoves]

But everyone seems to like Sun and its trendy to hate Oracle. Give them someone to rag on that's not MS or apple.

Re:Blame Sun, not Oracle

[gtall]

Yes, and the glove slipped so easily over Oracle's hand that they didn't even notice.

Re:Oracle is Evil, C# Java

[ADRA]

Harmony classpath are actually not too bad. I don't know comparatively how they stack up against Hotspot's classpath, but it seems to be working for android anyways. The lacking member of the equation is GCJ or any other clean room JVM implementations which sadly do lag behind the official releases.

Re:Oracle is Evil, C# Java

[Dog-Cow]

Getters are good because they abstract the data model away from the object interface. It doesn't matter where the data comes from or how it's stored because access is always through a method (if needed).

What's stupid about Java is that it doesn't hide the getters and setters behind properties. Just like the data model should be irrelevant, so should the fact that you may be calling a method to get a value.

Object Pascal (and perhaps C#) does it right. The getter may be a private data field or it may be a method. It's unimportant to the user of the object.

Re:Oracle is Evil, C# Java

[randallman]

The argument isn't about the license, it's about patents. At any time, MS could step in and say that they own patents essential to the Mono implementation. That's why people are cautious about Mono and why most steer clear of it.

Re:Oracle is Evil, C# Java

[BitZtream]

The exact same thing applies to Java as well then.

Re:Oracle is Evil, C# Java

[msclrhd]

Parrot? That is a VM that can run a lot of different languages. You could always take one of the JavaScript engines -- V8, TraceMonkey+JaegerMonkey, JSC, etc. -- and adapt it to run python if you were so inclined. Also, if you like C# as a language, you could use Vala. And fossils C and C++ may be, but a lot of software is built with them including the major OSs, Web Browsers, Compilers and Virtual Machines/JIT engines.

Translation

[SnarfQuest]

Apache Declares War On Oracle Over Java

Slashdot never definess anything. Translated, this reads

An American Indian declares war on an ancient Greek prophet over coffee.

Re:Oracle is Evil, C# Java

[GWBasic]

Mono should be looked at like WINE, useful to port programs to, useful to get some programs to run, but shouldn't be your language of choice if you want to get cross-platform apps.

I write ObjectCloud in C#, test on Mac with Mono, and deploy in on Ubuntu Linux with Mono. My experience with Mono is that it's fast and reliable, as long as you're sticking with the lower-level CLR APIs. IE, it's fine for servers that handle their own sockets; but it's not good for GUI applications.

Re:Oracle is Evil, C# Java

[aristotle-dude]

GP: Mono is not portable
P: Yes it is! Here is a link to its license!

Being open source doesn't make something portable.

Absolutely correct. I have tried in vain to get Monodevelop working fully on OS X but to no avail. There are a bunch of linux specific dependencies required to have it work fully. You cannot build most of the templates on OS X let alone being able to edit a GUI inside of Monodevelop.

The current state of the OS X port is an absolute joke and show how much linux is trying to copy the "windows" way of doing things.

FFS, this is bad...

[JAlexoi]

Any Java developer worth their salt, will know that anything else coming out about Oracle's plans for Java are nothing compared to this. ASF is probably the biggest source of software for Java developers. To the point that most Java software has components from ASF bundled, even if indirectly.
All of Oracle's Java based software has components from Apache. IBM's Webshpere software has components from Apache. JBoss, Spring, Google's tools... All of them...

Re:FFS, this is bad...

[fishbowl]

A lot of the people who comment on the subject seem to be familiar with the Java language itself, but not so much with the significance of the frameworks and libraries that are out there. In these threads, I don't usually get the sense that some of the posters are very aware of just how much business software has been built in Java in the past decade. Whenever I see comments dismissing Java based on stuff like applet or Swing performance, it just drives the point home that some people simply don't understand where the Java code is. (Hint, it's not in the end-user GUI or the 2D or 3D animation.)

Re:FFS, this is bad...

[steelfood]

All you really have to do is bring up task manager or run top and look for the biggest memory hogs. 95% of software written in Java will suck up memory, especially after long periods of use. The only way I've seen people (Oracle) get around this is to spawn new processes to do all of the heavy lifting.

Re:Oracle is Evil, C# Java

[jd]

Most of the extensions in C# were done better in D. And most aren't really necessary or useful. Now, this isn't to say I think Java is at the pinnacle of software engineering - it doesn't support a number of very nice features in "pure" OO, and there are "newer" concepts in Software Engineering (such as Templates, Aspects and Mobile Threads) that have no direct parallel. Obviously, some of these aren't getting used anywhere else either, so clearly they're not considered vital by anyone. At the moment, at least. In some respects, adding new concepts can be a bad thing, but it does mean that when those concepts are valuable, languages too rigid in their design won't be able to make efficient use of them. (C++ tries to include all kinds of new concepts, which is why it is so horribly complex. This is the price of being too fluid.)

"Better" is therefore hard to define. Better for what? The only real definition of "better" in any abstract sense is that the better language will be capable of evolving in the right direction at the right time without having too much baggage. And this can only ever be known in hindsight. Which is why Fortran is still useful in mathematics, even though it is a ghastly language -- it has evolved in ways mathematicians/engineers needed more than rival languages. In a battle of "survival of the fittest", it has survived and has thus demonstrated that it is fitter for some purposes. The usefulness of Fortran is utterly non-obvious and is only knowable because it exists.

In principle, on SMP/multi-core machines, Silk (an instruction-level parallel C), UPC and OpenMP-enabled C should have overwhelmed vanilla C by now. In practice, people just moved to better fork management, better threading libraries and better shared memory and used that to handle the parallelism. (The ATLAS maths library recently stopped developing its OpenMP support because it was inferior in practice.)

Re:So?

[Anonymous Coward]

>I have plenty of alternatives.

Ok. I have a mature product, which is a custom manufacturing ERP system. I am quite proud of it, it represents the peak of my career. It has been built in J2EE, and has some components deployed in JBoss, and some in regular Tomcat instances. A small ($50 million/year) company runs its factory on this system, which is responsible for supply chain, procurement, inventory control, and cost accounting. The system is dependent on more than a handful of items from the Apache toolchain.

In my shoes, would you be able to explain to my boss, how "plenty of alternatives" fit into this scenario?

Re:Oracle is Evil, C# Java

[fnj]

The current state of OS X ports of _MOST_ software is an "absolute joke."

Re:Oracle is Evil, C# Java

[gbjbaanb]

they're rewriting Hudson in C? really? show me a link (I'm actually genuinely interested now :) )

Wait for it...

[LynnwoodRooster]

Time for Oracle to circle the wagons and call for the cavalry!

Good job, Oracle

[Arancaytar]

- Apache Software Foundation: Check
- OpenOffice.org/Open Document Foundation: Check
- MySQL: Check

At this rate, you'll have pissed off the entire world of free software before the year is over. Maybe go for Linux next. Or the Mozilla Foundation, but I don't remember if Sun was involved there in a major way.

Re:Good job, Oracle

[vegiVamp]

I'm a MySQL dba, and I can't say that I'm particularly pissed off. If anything, it looks like *more* releases are being done, *more* bugs are getting fixed and *more* features added.

This may or may not be the impact of what Sun was doing internally before the takeover, but if they were gonna harm MySQL I'd expect them to hamper that flow. We'll find out soon enough, I guess.

Re:Good job, Oracle

[magus_melchior]

Oh, that's easy-- patent infringement lawsuit on both the Linux kernel dev team and Mozilla. Or if they want to hit absolutely everyone in the OSS/Free Software world, they'd file a patent infringement lawsuit against the FSF.

Of course, that would be like firing the sawdust cannon-- only instead of sawdust, you're firing thermite. Larry might be power-hungry enough to do it, I don't know (given the protracted cost of these lawsuits, it's usually not about the money, but about how much of the market you can deny the defendant).

Re:Good job, Oracle

[t2t10]

"Open source" Java developers have never cared about this stuff. They were Sun's lapdogs, and they will be Oracle's lapdogs.

Re:Oracle is Evil, C# Java

[gbjbaanb]

lol. Good job you don't use C!

(BTW, C# is a shoot-yourself language now - ever since they added var types, extension methods (omg yuk!) and similar stuff to support scripting languages built on the CLR. You can be spectacularly stupid without even realising it. They started well, but then.. kept adding stuff they liked. It'll end up like COM did).

Re:Oracle is Evil, C# Java

[gbjbaanb]

except that in C#, there's a handy shortcut that's used a great deal.

public type x { get; set; }

which provides you with a getter and a setter method on that type without having to bother typing all that nasty code to prevent invalid values from being set. In fact, they kinda provide you with all the style, but none of the substance of get/set methods in the first place, which I find to be pretty useless. You might as well access the member variable directly. I mean, for all the good it does. It does stop your StyleCop checker from telling you off for not using get/set accessor methods though, so that's quite worthwhile. (unless you think that automated style police tools are just part of the pointless bureaucracy too)

That, I think, sums up a lot of C# - nice IDE, some nice easy-to-use features, but ultimately, its just a very verbose language that doesn't really help you write good code.

Re:Oracle is Evil, C# Java

[assassinator42]

C# has unsigned data types. The lacks of which has always bugged me about Java.

Do not underestimate the importance of ASF

[brunes69]

It is interesting because from the point of view of any developer using Java for real work(tm), the ASF is just as important, if not more important than, Oracle at this point. I have the choice of several JVMs, but there is no real alternative to the multitude of Apache core libraries and frameworks.

Re:The Oracle at Delphi, Indigenous Tribe, Islands

[sznupi]

Sounds fine, as long as galleons will lose the ability to sink nuclear submarines.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

excuse me, but developing a C# application against the mono framework is completely portable.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

MS has a legaly binding statement that says mono and anyone who implements it is not liable under any MS patents. Your argument is an ignorant non-starter.

Re:Oracle is Evil, C# Java

[TrancePhreak]

C# is an ECMA standard. It would be pretty difficult for MS to sue the mono project.

Re:its on s60 as well

[PCM2]

Maybe not. According to the Wall Street Journal , RIM is planning to move all of its devices toward a new OS based on QNX, beginning with the BlackBerry tablet. No word on what that might mean for Java development on the BlackBerry platform.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

why isn't it good for GUI apps? GTK# is a good cross platform toolkit and has a good designer in mono develop.

Re:Oracle is Evil, C# Java

[exomondo]

C# is better than Java anyway

Citation needed.

Posting that is idiotic, if you know anything at all about those languages - or programming in general - you would know that there is nothing he/she can cite that would prove that statement one way or the other. Neither can be defined as better and that statement - if not accompanied by a specific situation in which one language is preferable - always comes with an implied 'IMO'.

Re:Oracle is Evil, C# Java

[jrumney]

C# is a shoot-yourself language now - ever since they added var types, extension methods (omg yuk!) and similar stuff to support scripting languages built on the CLR.

Even at the start, when C# allowed exceptions to be thrown without declaration, it had already started down the route of shooting yourself in the foot for the type of large software engineering projects with many developers that Java is mostly used for.

Really could be a trap

[williamhb]

The principle of estoppel would seem to apply here however:

Microsoft has promised not to sue, and stated so publicly, in writing. The Mono developers (and users) have proceeded under the assumption that commitment was made in good faith; Even if Microsoft reverses their decision, they cannot then sue for infringement of the *patents they already agreed not to sue* over. Estoppel would kick in, protecting the devs & users.

There is a bigger problem: Oracle again! We know that Microsoft, for .NET, licenses certain patents from Oracle (and that Oracle earn quite a bit of money from it). Does Mono have a licence to those Oracle patents? If not, regardless of what Microsoft say, what is to stop Oracle from suing the pants off Mono (or companies that use it) for patent infringement as well?

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

How is using a current version of mono + GTK# different than me using .net 3.0 + GTK#?

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

They have said they own the patents and they have publicly affirmed they will never sue mono or people who develop using mono. case closed... next!

Re:Oracle is Evil, C# Java

[Steffan]

Actually, I believe that 'Mono' comes from the word for 'Monkey'

3/10

python is on by default

[Kludge]

Python is included in the distn by default. Java is not.
Developers can develop in C or Python without adding a run time to the system.

Re:python is on by default

[steelfood]

The point is, Nokia foresaw the strategic value of actually owning its own platform instead of using the trendy new technology, and did what it needed to achieve just that. This will just be incentive to move to fully using Maemo on most, if not all of their fully-featured phones.

Note that they acquired Trolltech a few years back right when they became serious about shifting to Qt.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

you have to go WAY out of your way to use var and extension methods. It is still much more natural to statically type your variables and define all your methods in the class definition (well... extension methods could be a bit of an issue since it is easy, but I hate dealing with them so I don't care)

Re:Oracle is Evil, C# Java

[sockman]

Enums in Java allow behavior to be added to the class itself. Even if the state is immutable, the ability to extend an enum to add complex behavior based on the enum type itself, is something lacking in C#'s C-like enum support.

Re:Oracle is Evil, C# Java

[CodeBuster]

Even RSM has pointed out on numerous occasions that free software is not immune to software patents. Indeed, he rarely misses an opportunity to remind everyone of this fact. I am very tired of hearing free software advocates bring up the patent issue and then act like free software is not equally or even more encumbered by these patents; software patents are a threat to all software, regardless of source.

Re:Oracle is Evil, C# Java

[sockman]

I love being idiotic! "C# is better than Java anyway" is a blanket statement from an subjective zealot. Both languages have good features and bad features, but simply dismissing one because it isn't in your camp is rather short sighted.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

verbose? you must write C code.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

or... you know... not

http://dotneverland.blogspot.com/2008/07/yet-another-benchmark-of-clr-vs-jvm.html

Re:The Oracle at Delphi, Indigenous Tribe, Islands

[Kaeso]

FTFY.

The Apache Indian in North America would go to war with the Athenian Oracle at Delphi over the island of Java in the South Pacific.

Phocian, not Athenian... We mustn't let that Apache get lost while he's in Greece.

Re:Oracle is Evil, C# Java

[DougBTX]

Using var in C# still statically types your variables, perhaps you are confusing var with dynamic. I don't understand what you mean by "out of your way"... if you're using resharper, it will give you a handy prompt to convert redundant type declarations to use var.

Re:Oracle is Evil, C# Java

[Anonymous Coward]

And if you believe the statement is actually legally binding (did you ask a lawyer? Thought not), and if you believe there are no loopholes in it, then you are what we call a "convenient idiot". Don't be offended, it's a technical term. Look it up.

Google thought Java had a legally binding statement that said anyone who used it was not liable under Sun patents. Look where that got them.

Lava?

[turkeyfish]

So what happens when Apache uses the GLPv2 version of the Java LANGUAGE and builds its self a VM (not a JVM) called Lava that is attractive to Open Source Developers outside the JCP? Even IBM could get leverage on Oracle if they decided to cooperate. Oracle may own the JVM but it doesn't own the VM concept, so others have a lot of flexibility. At that point, it just requires the larger community to pile on and decide they want to use it. Java becomes history covered by Hot Lava.

Every one keeps saying Oracle is going to take Google down, but if Google can show it is only using VM techniques rather than JVM techniques, I can't see how they are not in the clear. There is just way to much prior art (UCSD-Pascal 1970's vintage).

Re:Lava?

[shutdown -p now]

So what happens when Apache uses the GLPv2 version of the Java LANGUAGE and builds its self a VM (not a JVM) called Lava that is attractive to Open Source Developers outside the JCP?

If they build a VM that is fast enough to be competitive, it's highly likely that they'll run afoul one of those pesky Oracle patents. After all, what you describe is precisely what Google did with Dalvik, and they got sued over it.

Re:Oracle is Evil, C# Java

[SplashMyBandit]

Mono is the *incomplete* open source implementation of .NET. There, fixed it for ya. Dontcha know that Mono doesn't implement all the required libraries for true interoperability with .NET (the Mono team*are* working on this, but let's get real about where the state of play is).

Re:Oracle is Evil, C# Java

[GWBasic]

When I said GUI, I meant WinForms. I should have been more specific. A WinForms C# App isn't going to run well on Linux. I have no experience with GTK#.

Re:Oracle is Evil, C# Java

[the-matt-mobile]

...and useful to extend to pool of developers that can write programs for the Linux platform. Developers are the only way to get and keep users. That seems to be the one meme that Microsoft has always gotten right, and the place where the Linux community could use some serious self evaluation.

oracle business depends on open-source.

[MadMaverick9]

The Oracle HTTP server is a simple Web HTTPD server (Web listener). It is based on the Apache Web Server provided by the Apache Group

http://www.orafaq.com/wiki/HTTP_Server_FAQ

Oracle Enterprise Linux (OEL) is Oracle's Linux distribution. OEL is based on Red Hat Enterprise Linux and is binary compatible with Red Hat (not a fork!).

http://www.orafaq.com/wiki/Oracle_Enterprise_Linux

So why isn't the open-source community taking advantage of this leverage? Just prevent Oracle from using your products for free. Make sure Oracle pays you (you = Red Hat, ASF) when they download your products. Oracle is vulnerable. Take advantage of it, for crying out loud.

Oracle is using your stuff for free, Mr. Red Hat and Mr. ASF, and making big money with it.

Re:Oracle is Evil, C# Java

[shutdown -p now]

There is still one that FreeBSD has (that was actually "blessed" by Sun).

If I remember correctly, FreeBSD's version is a port of the original Sun implementation, not a grounds-up rewrite. So any patents would still apply.

Actually, given how they're suing Dalvik, which is not only a complete rewrite code-wise, but is not even a Java VM - I think that, for the time being, the only safe assumption is that any VM is in the patent mine field. You want to be safe, stick to languages with implementations which compile directly to native code - C, C++; if you want something more high-level, Vala.

Re:Oracle is Evil, C# Java

[shutdown -p now]

For instance, Java has much better enum support (they are proper objects) from my perspective.

There are precisely two things that Java as a language has over C#:

- better enums
- "super" in generic wildcards

On the other hand, here's what C# - as of version 4.0, the current stable one - has over Java:

- first-class properties
- delegates & events (yeah, you can do the same with interfaces and anonymous inner objects, but it's 4 times as long)
- full-featured closures
- type inference for locals and lambda arguments
- user-defined value types
- nullable value types (with no boxing)
- ability to extend classes and interfaces from the outside
- operator overloading
- RAII blocks ("using")
- sequence comprehensions with syntactic sugar (LINQ)
- dynamic name-based member and operator dispatch ("duck typing")
- C-style pointers with pointer arithmetic and unions (for when you want to get closer to the metal)
- ability to directly call any C library (P/Invoke), with a type system that fully covers everything in C with direct mappings

I've probably forgot a few. From this list, RAII will appear in Java 7 (IIRC), and closures only in Java 8, and that's about it. Of course, by that time we'll see C# 5.0 with syntactic sugar for asynchronous code.

Making the whole world your enemy?

[JSBiff]

I'm confused. Nobody can succeed with the whole world as your enemy. Maybe I'm missing something, but from where I sit, it doesn't look like Oracle has many friends to begin with, and is losing more every day. Oracle and Microsoft have basically been enemies forever, but Oracle had other companies which were it's friends, of sorts, like Sun, IBM, and some others. It just looks to me that Oracle is going to make everyone into their enemy, and how can any company succeed, long term, with that situation?

Re:Oracle is Evil, C# Java

[shutdown -p now]

BTW, C# is a shoot-yourself language now - ever since they added var types, extension methods (omg yuk!) and similar stuff to support scripting languages built on the CLR

You are thoroughly confused.

There is no such thing as a "var type" in C#. When you see this:

var x = 123;

it means "declare a variable named 'x', and infer its type from the initializer". In this case, since the initializer is an int, so shall be the variable. It doesn't have a "var type", and it cannot change type later.

Extension methods are there, but they have nothing whatsoever to do with "supporting scripting languages". They appeared in C# 3.0 to back the syntactic sugar in LINQ.

There is precisely one feature in C#, which appeared in C# 4.0 (i.e. 8 months ago), that is there to "support dynamic languages", and that is the "dynamic" type, as in:

void Foo(dynamic x) { // duck typing
  Console.WriteLine(x.Length); // Length is looked up at runtime by name
}
x("Foo"); // okay, will use String.Length
x(new[] { 1, 2, 3 }); // okay, will use Array.Length

which is most likely what you have confused with "var". You are, of course, free to use it or ignore it. So far I've only seen it once in the code which has to work with MS Office via COM, and the code would have been much more painful if not for it.

Re:Oracle is Evil, C# Java

[shutdown -p now]

Even at the start, when C# allowed exceptions to be thrown without declaration, it had already started down the route of shooting yourself in the foot for the type of large software engineering projects with many developers that Java is mostly used for.

Of all languages with exceptions, Java is the only one I know of which has checked exceptions. Hm, wonder why...

Also got to wonder why the single most popular Java framework deliberately avoids checked exceptions:

Checked exceptions are overused in Java. A platform shouldn't force you to catch exceptions you're unlikely to be able to recover from.

Re:Oracle is Evil, C# Java

[shutdown -p now]

by and large the rest of that superset is complete crap - most of it is just a shortcut to bad software engineering.

It would be helpful to be more specific - which features exactly are a "shortcut to bad engineering", and why? Are you talking about first-class functions (delegates)? Closures? Sequence comprehensions (LINQ)? And if they're so bad, then why Java keeps trying to chase C# on all of those?

Re:Oracle is Evil, C# Java

[shutdown -p now]

To be honest, about the only case where unsigned data types are actually used in C# is when you're calling into some C or COM library. Outside of it, it quickly becomes a mess once you start mixing signed and unsigned values, same as it always was in C.

The issue at hand?

[slack_justyb]

Some people are wondering why Apache just won't align itself with Oracle and be done with this.

It's not about the GPL vs Apache License.

The true question is the same that faces Mono and .NET.

Is the Java Platform an open standard that anyone can implement or is it an implementation of a standard that Sun/Oracle made up? We all know that in the .NET world that Mono is just an implementation of a Microsoft standard that later became an ECMA (some parts) standard. Apache wants their Harmony project to be under their license not the GPL.

Let's look at this from say a C or C++ view point. Could Apache write a C/C++ compiler and library and put it under the Apache License? The answer is yes, there are a couple of patents out there that they'd have to watch for but without hesitation it is in the realm of most absolutely. Now turn that same logic over to say Java or .NET. You'd find yourself in the realm of Not really or Fat chance.

So in the end is Java truly open or are we just open sourcing something that is controlled by one entity? If you look hard at the OpenSolaris and the OpenOffice.org debates you'll see that there is a common theme going around here. Me thinks that Apache is grandstanding for a point and hoping the current tide (the anti-Oracle movement) helps push Java somewhere, where the person who is truly in control won't let it go in or at least won't care if ASF decides to leave.

all this over...

[celle]

An indian declares war on a priest over coffee. What happened? Someone spill the cream? Lid fall off the sugar? Get over it already! Damn oversensitive babies!

Re:Oracle is Evil, C# Java

[jrumney]

Checked exceptions are overused in Java. A platform shouldn't force you to catch exceptions you're unlikely to be able to recover from.

Java doesn't force you to do that. If the caller is unlikely to be able to recover, then you should be throwing a RuntimeException or Error, both of which are unchecked. But for exceptions that can easily be handled, it is much easier to handle them when you first write the program rather than have to go back and find all the problems later when your testing shows up that you've missed some exceptional conditions that commonly occur.

Re:Oracle is Evil, C# Java

[shutdown -p now]

Java doesn't force you to do that. If the caller is unlikely to be able to recover, then you should be throwing a RuntimeException or Error, both of which are unchecked.

Yes, and that is precisely what Spring does for most of their exceptions.

The icarusian tale of oracle

[jonaskoelker]

If their talented engineers are by-and-large leaving and they are by-and-large unable to hire more, they will quickly become a dying shell of a mediocre company.

So you're saying their fall started when they got too close to Sun? Sound like Icarus ;-)

Existential issue

[aplcomp]

Life is too short for Java

Re:Oracle is Evil, C# Java

[Migala77]

enums are not objects in Java, though they must be declared within an object. Which is stupid, as enums are types, just like objects are.

No, enum types are classes, and enums are objects. They derive from Object (via Enum) and can be declared in their own file, seperately from all other classes. The only difference is enums have a limited set of instances.

Re:Oracle is Evil, C# Java

[TheRaven64]

Or you could, you know, compile your code to native machine code. Seriously, when was the last time you actually needed the same binary to work on more than one platform? If you one of the tiny subset of the population for whom the answer is not 'never,' when was the last time that this couldn't have been solved with application bundles containing multiple binaries?

good riddance?

[Bouncelot]

I hope this is the end for Java. That language never lived up to its promises, and it it wasn't for the media hype, and the dumbing down of the software industry, it never would have taken off in the first place. I'd gladly trade 10 java devs for a single C++ dev.

Re:Oracle is Evil, C# Java

[TheRaven64]

Java doesn't force you to catch exceptions that you're unable to recover from, it forces you to either catch them or declare that you propagate them. This is incredibly important if you want to do static code flow analysis, both for optimisation and for validation. One of the things I'm currently being paid to work on is exception support in C++, and the lack of this information in most C++ programs makes life very difficult for the compiler.

Re:Oracle is Evil, C# Java

[TheRaven64]

The point of using set and get methods like this is so that you avoid exposing the implementation with the interface. It's irrelevant in a language like C++, where the header needs to contain the class layout, but in a slightly more dynamic language it's very important. In this version of the class, the set and get operations are just writing to the instance variable directly, but in a future version they might do something more complex.

I'm not sure about C#, but in Objective-C they also allow things like key-value observing to work automatically (i.e. no user code required). The framework swizzles the set methods at run time, replacing them with one that calls the original bracketed with calls that deliver notifications. This allows for some very loose binding between models and views, with entirely generic controllers. If you access the ivars directly, there is nowhere for the framework to insert the hooks.

Re:The Oracle at Delphi, Indigenous Tribe, Islands

[shikaisi]

Well, if you saw them, sir, they weren't Apaches.

Re:Oracle is Evil, C# Java

[VGPowerlord]

That shortcut actually exists for a reason.

To be specific, if you later decide you want to implement logic while getting and setting, if you made it a public class variable instead, you'd have to recompile everything that links against it.

If I put public type x { get; set; } and later decided that I wanted to make sure x.foo != 0 during the set, I could change it to something like this:

private type myX;

public type x {
        get {
                return x;
        }
        set {
                if (x.foo == 0)
                      throw new ArgumentException("foo cannot equal 0");
                myX = value;
        }
}

This requires no recompilation of any other classes, just of the class this logic is in. Granted, it's still a bad idea, as callers wouldn't be expecting an ArgumentException, but it makes sense if my logic was going to divide by that value at a later date... I've just shifted it from a DivideByZeroException to an ArgumentException.

Re:Oracle is Evil, C# Java

[mcgrew]

I wouldn't consider six years to be be "shortly."

You will when you're my age.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

Well... that would be like saying the KDE framework isn't good for GUIs since it doesn't support win32 or winforms UIs.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

you are right... I was thinking implicit typing versus explicit typing.

I still do not use it and since nothing forces you to use it, you have to go out of your way. I feel it was put there to make VB developers feel more comfortable.

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

I was more interested in the content than the actual pictures... since the conclusion was basically

"these numbers mean squat since the performance variance depends greatly on the underlying environment"

Reading must be a dying art.

Re:Oracle is Evil, C# Java

[GWBasic]

No, that's understanding the design of the CLR. It allows APIs that are platform specific and APIs that are platform neutral. Assuming that you build against Microsoft-specific or Windows-specific APIs, then you're not going to run well, or at all, on Linux.

Microsoft might be Evil but Oracle is the DEVIL

[TiPros]

Microsoft might be Evil but Oracle is the DEVIL

Re:Oracle is Evil, C# Java

[modmans2ndcoming]

I like to make it easy to read my code.

I don't care what MS does in their libraries, but my code will be explicit. I even use nice long variable names with underscores between words in the name!

Re:Oracle is Evil, C# Java

[Tablizer]

And fossils C and C++ may be, but a lot of software is built with them including the major OSs, Web Browsers, Compilers and Virtual Machines/JIT engines.

It's a matter of which niche they better fit. C/C++ is best when speed and efficiency matter most, such as system (base) tools. Custom Application development on the other hand can accept some inefficiency in place of expressiveness and dynamism.

Re:Oracle is Evil, C# Java

[t2t10]

C# is tied to a single OS. That makes it a nonstarter right there.

Runs fine on my Linux box.

Mono is not a portable version, it is like its namesake a disease. Meant to poison the well that is Free Software.

Funny, given Sun's/Oracle's patent pool and litigiousness, I think it's quite clear that Java is "meant to poison the well that is Free Software".

Re:Oracle is Evil, C# Java

[t2t10]

Do you not think it exists solely to get MS patents into the free software ecosystem?

And which patents would that be? Nobody has been able to show a single essential patent necessary for implementing core C#, and yet Microsoft also made a legally binding commitment.

For Java, not only are there dozens of essential patents, we've known about them for years, and Sun/Oracle explicitly doesn't license them. And Sun/Oracle has successfully poisoned the open source ecosystem with their fake open source software.

Do you think MS is just going to let it thrive ever?

Both Oracle and Microsoft are totally untrustworthy and evil; however, if I have to deal with evil companies, I prefer to do it with some legal protections, and Microsoft offers those while Oracle doesn't.

Re:Oracle is Evil, C# Java

[samwhite_y]

There is another big difference between C# and Java. In Java, you are strongly discouraged from making native calls. For example, instead of using the native desktop GUI widgets, Java writes a large wrapper (Swing) around the desktop interface and tries to port all that complexity from platform to platform. In C# on Windows, I make Windows (C# friendly wrappers -- but still Windows) specific calls if I am trying to create a GUI user interface. Similar points go for certain other APIs such as ADSI, network pipes, registry, HTTP, and encryption. In fact, if you look at the general low level Windows APIs, there is a lot of functionality there that is not captured in Java.

I agree that C# is better than Java at doing the basics, but C# is hard to separate from the platform that birthed it. This makes C# a difficult language to deal with if you are not writing for a Windows platform. For example, from my understanding, the port to Linux of C# is not to make applications portable from Windows to Linux (except maybe some server apps), but to make C# a productive language for Linux the ways it is for Microsoft. I expect to call native Linux APIs from C# when I am running on Linux. Why not? The benefit of doing otherwise is not so clear given the general lack of portability of C#. This approach does create stress points. The .net API is a large collection of APIs, some of which run on Linux on C#, some that don't. Some are protected by GPL like licenses so you can use them without a fear of a lawsuit, others are not. As an example, I think there are still some controversies about some of the fancier parts of the WebForms APIs (some of the complicated dynamic HTML tables for example).

Re:Oracle is Evil, C# Java

[Samus]

I used it earlier this year for writing a library. While it wasn't perfect it was passable. It may have crashed a few times but I never lost my work.