IntelliJ Now Open Source

JetBrains just announced an open source edition of their popular IntelliJ IDE, under the Apache 2.0 license.  As many might know, IntelliJ was one of the last IDE’s that actually cost money.  It appears they have adopted an open source plus commerical value add strategy, similar to many companies in the Eclipse ecosystem.

This announcement is pretty surprising; at least to me.   It will certainly make it easier to get IntelliJ and hence competitive to Eclipse IDE.  However, it might open up opportunities for the two communities to collaborate?  What I would love to see is IntelliJ provide great support for Equinox and OSGi; more and better tools make the world a better place. 🙂

Congrats to the JetBrains team for making the big leap.  I am positive your decision to go open source will pay off for you in the long-term.

8 Responses to IntelliJ Now Open Source

  1. Mike P says:

    “It appears they have adopted an open source plus commercial value add strategy, similar to many companies in the Eclipse ecosystem”

    What does this strategy look like. As in, how can a company still make an income, but also offer free software?

    -Mike

  2. Donald Smith says:

    @Mike P

    Just look at the IntelliJ website. They have the free, open source “community edition” and then they have a for-pay enterprise edition with complete J2EE support, more advanced tools, etc.

  3. Hi Ian,

    I’d disagree with your interpretation of this JetBrains’ move. Let’s guess their motivation:

    Please do not forget JetBrains is a small company, just probably OTI size or something like that. So a few hundred more users paid for IDE is probably nothing for IBM but a value for JetBrains. Moving to open source will very likely decrease their sales in short term because some potential users who are aware of IDEA (like me) will use free one. There will be some time while open-source will spread a world and reach potential users who are not aware of IDEA (if such people ever exists). And some of their potential users who are thinking about to buy IDEA will be OK with free version.

    So considering this fact let’s think why they decide to move open source?

    Guess #1: Their are focusing on more profitable products like ReShaper; do not count on IDEA as a profitable product and let it be open source to gain indirect benefits.

    If this is true, nothing positive about this move, and this is just a beginning of IDEA end.

    Personally I do not believe in case #1

    Guess #2: This is beginning of world-wide “JetBrains attack” on the IDE market. There is only 2 products which were restraining such an attack: eclipse and netbeans. Because while eclipse and/or netbeans are strong — such attack would be very stupid move for (again!) small self-funded company. If guess #2 is closer to the reality, and JetBrains decided to open source IDEA — this is a serious message to eclipse: JetBrains do not consider eclipse as a competitor in the middle term, and they believe they can take significant part of current eclipse and NB market.

    Personally I more believe in guess #2, considering stalled innovation in the eclipse ecosystem (besides modeling) and inability to explain where eclipse is moving and low it will look like in 2010 or 2012…

    Reality of course is much more complex, but I can’t see any good signal for eclipse ecosystem in relation with this event. Collaboration between two communities? Nonsense for the similar reason: JetBrains is very small. This could be possible if JetBrains will have 50% of eclipse market, which just support guess #2. And of course if they even had such plans — they were able to make them real many times.

    Kind Regards,
    Andrey

  4. Ian Skerrett says:

    Andrey,

    You might be right with your guess #2 but how does JetBrains make money and pay for the “attack”?

    Probably not surprisingly, I disagree that there is a lack of innovation in Eclipse ecosystem. Mylyn is a great example, MAT is another in the tooling space. Lots of things going on in Eclipse runtimes. As you mentioned modeling has lots of new stuff. Also if you look at the mobile space or cloud computing, most of the new tools coming out are Eclipse based. I am very happy with the level of innovation in the Eclipse ecosystem.

    I actually have a lot of respect for the folks at JetBrains. I know lots of people like their product and I know it is a challenge to work in a small self-funded company. Am I worried that JetBrains is going to take over the IDE market, no not really. If that is their goal, good luck to them.

  5. Hi Ian,

    Thank you very much for reply, and please understand my motivation for the post. Eclipse was a greatest technology I met in my life. And if my words sounds negative — that do not mean I’m negative about eclipse technologies. I can’t be, especially all projects I’m working on since 2002 up to now are eclipse based. I’m just really really sad about eclipse is going down for past year, this is very very sad to me but true.

    I can post in details fact about stalled innovations in every area you mentioned (except modeling), if you are interested.

    As a short response: MAT is cool, but this is just a tool — very good tool solving plain old practical need, there is nothing innovative in heap walkers since past century. Mylyn? Is it as much innovative as bugzilla? BTW it’s great to know that p2 buzz is finished.

    New tools for mobile, cloud and other spaces are coming, and there will be much more in short term, but this is wrong indicator of innovations at eclipse, let me please give you my example:

    Within our company we’re thinking about technology stack for next ten years or so. (BTW eclipse RT and technologies is our current stack since 2002). At the moment we have a vision for such stack, and fundamental decision we need to made is about programming language to use in this stack… This will not be Java but widely unknown Fan programming language: http://www.fandev.org. Fan is really cool language among many other innovative languages on JVM and CLR, which born recently like Clojure and Microsoft’s F#… can’t forget Scala here🙂

    All of these languages still lack of support from tooling side, runtime side, etc… so we can see many initiatives to build tools for this new languages and technologies. For example: Scala in OSGi, eclipse-based Scala and Clojure IDEs, and we’re doing very similar things for Fan: running Fan in Equinox container and building eclipse-based Fan IDE. Is this innovation? Very likely. Is this innovation @eclipse? No!!!

    Let me please explain on my Fan example: without serious shift in eclipse ecosystem, eclipse do not look good for us as technology stack for next decade, and we’d be happy to have tools based on our new stack including IDE built on new technologies. This is very huge task, especially Fan is lacking of even basic UI library and of course lack of basic development tools. Instead of going this very long way, we are going to use eclipse to bootstrap process: 1) build eclipse-based IDE to support our development needs 2) use eclipse-based tool to build Fan framework, IDE, and switch to this new Fan IDE developed in Fan. (Honestly we do not plan to build Fan IDE in Fan because of limited resources, but I believe this will be done eventually by someone else). And like Scala guys we’re running new languages within OSGi, but not to innovate — just to technically support bootstrapping plan.

    So we’re considering Eclipse as a “boot code” to bootstrap development process with new languages and technologies. I believe when infrastructure (comparable to Java and eclipse) at new emerging languages will be relatively developed — all the developers will switch to tools, runtimes, and frameworks developed in that languages — just because new languages has features aimed to build better frameworks (in terms of concurrency, powerful type systems, functional programming, etc). Eclipse can’t compete with elegancy and power of emerging frameworks until it’s Java-based. OSGi will not be able to compete because formally OSGi is yet another Java-based framework…

    So there is definitely a lot of innovations in software engineering now, and world is changing very fast. Many of this innovative things rely on eclipse as tooling or runtime platform, and this will be continued until new technologies will be self-hosted and will produce own set tools/frameworks. This is a question of couple years. After couple of years eclipse/Java will look as innovative as emacs/LISP nowadays.

    Just an offtopic question: “did OTI use Visual Age to build initial version of eclipse?” If my assumption is correct, what happened with Visual Age after eclipse developers switched to eclipse?

    Example above is related to programming languages and Java language at eclipse (I know about support of JVM scripting languages @e4, but that’s another terrible story). Similarly no innovations at other eclipse areas, or just things, which looks like innovation, but not an innovation in fact (for example Mylyn will be alive until eclipse is alive, and will die as soon as eclipse — this is not true for MAT, which is self sufficient tool and will be useful for years independently of eclipse as a platform), or just cynical eclipse usage as a bootstrap for real innovation… just a form of fire-and-forget, and that’s really sad.

  6. Konstantin Scheglov says:

    I have to agree with Andrey, I see that Eclipse does not improve as much as it was in previous years. I look regularly on list of changes for each milestone and see that there are no reason to update. Just no any feature useful for me. In the past I’ve upgraded each milestone and often even to integration builds. But that was in the time when Eclipse was young and moved fast…

  7. Ian Skerrett says:

    @Konstantin,

    A lot of innovation at Eclipse happens in the other projects, not in the main platform. e4 is the place for the platform innovation.

    @Andrey

    Fan looks pretty cool. It will be interesting to see how it evolves. I think you under estimate the interesting things going on at Eclipse but that is okay. The exciting thing about innovation is you don’t know it is successful until well past the starting phase. The next five years will be exciting to watch; just like the last five.

  8. James Briant says:

    @Andrey

    I hope its #2!

    I’ve used eclipse for about a year, but I finally got fed up with its subversion integration and how that interacts with the maven project layout. I figured it was time to give IDEA a try, since I’ve used ReSharper for years. What a breath of fresh air. It just works. Its maven support just works. Everything just works. Even the beta releases of version 9 work.

    As for who would pay for the non-free version, that would be me. But then, I’ve donated to the eclipse foundation too. Those of us who earn a living through these excellent tools are willing to pay or give something for it.

    James

%d bloggers like this: