Vaadin licensed under Apache, but will it be changed to GPL

The strongest advantage of Vaadin is Apache license. It is the reason I started evaluating a couple weeks ago. Recently, I found Michael Widenius, a founder of MySQL, invested Vaadin (which is great). However, MySQL changed its license to GPL in 2000 to charge for commercial use (which heart me). Will Vaadin do the same thing after it gains the market share? I know it is a big ask but I am looking forward the promise from Vaadin’s founder. Thanks.

Hi Zodiac and thanks for a really good and forward looking question.

We (IT Mill) have internally discussed the licensing a lot and evaluated all the available options. As a result of this evaluation, we are happy with the selected Apache license and are not planning to change it in the future. We believe that the Apache license itself is the best guarantee the users that the product is truly free - today and in the future.

That said, who pays the bills? Our revenues are based on selling
add-on components
on top of Vaadin. Based on this revenue, we are able to continue investing in the
core product - Vaadin. All the commercial services and products on top of Vaadin are
100% optional

In the near future, we will open
a new channel called " Vaadin Directory"
for distributing add-on components and tools. Most of them are free. I would guess that over 90% will be licensed with a liberal open source licenses such as Apache or BSD. Some (optional) add-ons will be completely commercial or dual licensed.
is an example of a completely commercial add-on and
, to be released in few weeks, is an example of a dual licensed one. Note that use of these is not required by any means and 100%

exist for the both of them. Directory will also serve as a channel for you to either distribute your add-ons for free or to sell your add-on for a fee.

I hope I did convince you that Vaadin will be Apache licensed also in the future, we can keep investing in it in the future and is something you (and the rest of the Java community) are able to rely and invest your time in more easily than to any of the alternatives.

If you have any questions related to business, licensing or future - feel free to ask.

  • Joonas (founder and CEO of IT Mill, the company behind Vaadin)

Hi Joonas,

Thanks for the detailed and honest answer. It is really helpful.

It brings in another question when you say It Mill will make money from solutions and addon components. Could you provide more information about components and tools that won’t be free? I know it might be over-interpreted but it seems there is going to have something like Vaadin EE (or different name) in the future (like Z* EE and I* EE did). Will the Apache-ed Vaadin be then limited in functionally?


No - there will not be Vaadin EE (and thus “Vaadin Crippled Open Source Edition”).

Add-ons will be just that - add-ons. None of these is required for buildling great & large Vaadin applications. They just try to make the development even more easy and give additional rich widgets to show off in your application.

Some of the planned add-ons and their pricing (per developer):

  • Vaadin TestBench. 1199 USD / CVAL.
  • Vaadin TouchKit. 749 USD / CVAL. AGPL available for free.
  • Vaadin Timeline. 399 USD / CVAL. AGPL available for free.
  • Vaadin JPAContainer. 299 USD / CVAL. AGPL available for free.
  • Vaadin Calendar. 149 USD / CVAL. AGPL available for free.
  • Vaadin ColorPicker. 49 USD / CVAL. AGPL available for free.

(Disclaimer - these are plans and the pricing and availability is not set yet). Feel free to comment on pricing - all feedback would be great.

I hope that no-one will find it bad for making also commercials add-ons available. This allows us to increase our investments to Vaadin R&D, which benefits everyone. But we try to really careful in being fair and keeping Vaadin Framework useful, free and open source for everyone. In other words - want to ensure that the free Vaadin Framework will be the best available framework for building enterprise RIA - even without any commercial add-ons. The
commercial add-ons will just give even more options and make Vaadin platform even more interesting.

Furthermore - strong support for building and packaging add-ons opens up a channel for anyone - you - to publish and sell your add-ons too for Vaadin community. We believe that Vaadin Framework is a good basis for large growing community and ecosystem that benefits everyone involved.

EDIT: Removed a reference to ZK:s licensing model. My sincere apologies for anyone who found that reference to be insulting. The contents of the message has not been edited otherwise.

It is great to know Vaadin will stay free. My question is that the line between optional and required is blur and subjective. I know it is a bit push to ask your future plan, but it makes me nervous when you say “commercial addon components”. Let us take ExtJS as an example. Are they all ‘required for building great & large Vaadin applications?’ Or some of sophisticated ones are optional? Or, by optional you mean are application-specific components and advanced development tools?

I do not have crystal ball on future roadmaps, sorry. Thus it is hard to promise which add-ons will be created in 2011 or 2012.

( warning - business strategy discussions ahead :) )

From competitive point of view - IT Mill wants to keep/make Apache-licensed Vaadin Framework the best free framework for building RIA in Java. Because competitors will make their frameworks better, we are prepared to continue investing heavily in making the Apache-licensed Vaadin Framework better all the time - to keep it competitive. Freedom is one of the key values of Vaadin - we believe that the only way of gaining and maintaining community’s trust is to keep the framework truly free (both money and freedom to do whatever with it the community wants to). Only with this trusts people will continue to invest their time in and contribute back to Vaadin. Those contributions are crucial for making Vaadin the most popular RIA framework for Java.

That said - IT Mill is a corporation that tries to make money. We believe that if there is a healthy, happy and growing community and ecosystem around free Vaadin, it will allow us to offer add-ons, services and support for people who already trust and respect us. Having that trust and respect is good business and we intend to keep people happy :)

“Our revenues based on selling add-on components” is called dual-version. It makes no difference from zk, icefaces or smartgwt, no matter how you call it, or package it.

Vaadin and its supplements shall be 100% free – it is the strongest strategy. How to make money? Follow jboss and spring - sell the company at premium price. I don’t see any open source based company can really make money from license. just my two cents.

I must disagree. There is only one version of Vaadin Framework and it is released under Apache 2.0.

If IT Mill and others sell add-ons that support and add functionality to Vaadin Framework, it does not make the framework “dual-licensed” or even “less free”. In the same way Eclipse is not “dual-licensed” even though there are a lot of companies who offer commercial plugins that add functionality to Eclipse.

Per developer pricing is really weird for something like this. Just make it a flat fee that gives right to full support.

And why not LGPL – you are cutting yourself from all companies that design commercial offerings.

I personally don’t think that a component like a JPA container should be an add on. You want people to have a clear, supported, vendor independent model to handle persistence, not an incubator HbnContainer. Persistence should be in the base product. Then there could be add ons on top of a basic JPA container.

I have to agree, the license sounds weird. Are the licenses transferable to other people within the team, let’s say one developer leaves the team and another one joins replacing him?

And I have to agree on this point as well. The JPAContainer is one of the most requested features among the community (second most popular request on uservoice) - I don’t like the idea that you’ll have to pay for a feature that definitely should be in the core product. People have said that the lack of such a container has bean a deal breaker, the question is that will people accept the commercial solution, only time will tell.

Developer licenses are quite accepted as they are relative to the cost of development. While as flat fee would be useful, it is always hard to define the scope of the fee - is it one team - one project - one small company - one enterprise. Per developer is the most fine grained and thus allows to set the prices lower for small teams.

Sorry for confusing communications. The license for these add-ons is selectable: 1) use with AGPL for free, 2) use with “commercial vaadin add-one licence” (CVAL) for a fee or 3) do not use it. There would not be any incentive to select option 2 if option 1 would be LGPL, would there?

And just to re-state. We are not turning Vaadin commercial anyhow - instead we are increasing investment to Vaadin by starting to publish commercial add-ons in addition to what we would have developer in any case. Thus the availability of commercial add-ons should not take away anything from the community - only add choice.

That was a tough decision between starting a project that would produce a commercial JPAContainer or not starting that project.

There is already at least two free JPA implementations available for Vaadin: [list]




EclipseLink Container


Still we thought that JPA support could be better as many people wasn’t too happy with the above mentioned implementations. In order to be able to invest development time in creating something that would be better than what was already available we had to set a price for the results. IMO availability of commercial JPAContainer doesn’t anyhow take away the existing free alternatives or make them weaker. Also I hope that people would find Vaadin even nicer now when there is one more alternative for them.

Everyone is still free to make free JPA implementations better or start new ones; and I expect that they will evolve to match the quality of the commercial implementation some day in the future.

I think the central part to note here is that IT Mill and Vaadin are separate entities. Vaadin is free (as stated above) and in order to make Vaadin appeal to more developers Vaadin created the Add-on Directory.

IT Mill (who has heavy interest in Vaadin but is just one part of the Vaadin community) starts to create add-ons into the directory. These addons add value to the product and are worth paying for (that is the assumption). Also other companies are free to add their components to the directory in the same way. However, IT Mill still has the possibility to include its addons into the core product at some point if they see it fit and the R&D team of IT Mill is still focused on Vaadin as before eventhough some other parts of IT Mill create add-ons. I wouldn’t see this as a problem for Vaadin.

As for the fee, I have to agree with the flat fee. It is just so much easier to buy and already the fact that you have to calculate your cost for yourself raises the barrier of if you should buy it or not…“How many people work on our team actually? Does it include these guys who only work half-time? How about the testers? What do we define as a project? etc.” In this case I think ease-of-buying overweighs the benefits-based-pricing argument.

I dislike the idea that IT Mill will sale add-ons like Calendar, TouchKit, Timeline and so on … because they will never be integrated natively in the main Vaadin framework.

Developing applications and solutions by IT Mill to final customers is a good idea because IT Mill will find the same difficulties as us (developers) and will integrate solutions and enhancements natively in the framework.

But selling add-ons to developers gives the opposite result for us.

It’s just my opinion.

I would agree if ITMill sell only own addons. However, the idea of the Directory is to open an opportunity to any developer to distribute free and commercial addons, like apps in iTunes store. I think this would gain the community with the number of great components. The pricing politics, however, needs to be carefully designed.

Having only the limited/tightened framework and paid addons would be really bad, but this seems happily not our case. :slight_smile:

Speaking as a Vaadin developer, like yourself, and knowing in detail how the client-side part in Vaadin works, I must say this direction is currently much more fair to all Vaadin developers as a whole. Since each new component/widget will increase the loading time of the widgetset, I would be angry if the default widgetset would contain all the above mentioned widgets, which I have no use for. This could change later, though, if it would become easy for us to select which core components will be included in the final widgetset.

But because of the easy inclusion of new widgets straight from the Directory, I can still extend the core with the exact features I need. And I mean really, adding a new component is child’s play with Eclipse (hopefully we’ll get to help other IDE’s too soon). So basically this would mean exactly the same thing as selecting which core components will be included in my final widgetset.

And what I can say about the official add-ons, looking from inside the company, they will be as high standard as the core components (although not as mature, but some of the core components are just plain bad). What I’ve been told, they will be “certified”, by some of our staff. Anyone can ask for certification for their own add-ons, when the Directory opens.

And personally speaking, if I pay something for an add-on, I will expect to get some sort of guarantee or support for it, something which I will not expect from the free open source core. But I take no stanse in the pricing or license debate, since I’m really bad with numbers and money. I just expect the prices to be fair.

Yes, we’ve been doing this for the last seven years or so. And take my work for it, there have been difficulties :slight_smile: Most of the bugs and issues have been found and corrected because of the work that’s being done by our service department, using Vaadin to develop large enterprise level apps. The R’n’D department really doesn’t have any time to develop any actual applications, because all the time goes to fixing and making the framework better for everyone else.

And only the last year has really brought bug reports from the community and not customers, which is of course a very good sign.

…and why isn’t my Cameleon Theme add-on in Joonas’ list? And why will it be free, while the ColorPicker has a price tag? Sounds crazy…

Oh well, I guess my work effort isn’t as valuable as someone else’s :stuck_out_tongue:

Actually decisions on Chameleon Theme are not ready yet. We must discuss this early next week.

We are in process of adding possibility for setting volume discount steps. This work is not ready yet and will not make it to the Directory Launch. When ready, it would allow add-on makers to make the fee flat. Or to make couple of pricing steps (<5 dev, < 20 devs, <1000 devs for example). I hope that such feature can be added shortly and will make pricing much more flexible and address the concerns you outlined.

You outlined a good analogy. I’ll try reshape it in a way I see the picture:

The baker continues to make and give out bread for free. In addition to that he listens to the people eating the bread and adds salt with no extra cost. Some of the people get a great idea on how to make the bread more crunchy and they give a better recipe to the baker. (Submits a patch to As a result, the bread is better than ever and still free for everyone.

The baker also has a small restaurant connected to the bakery. He sells excellent meals (solutions, services) in there. While as the bread is the core of the meal - and the thing why most of the customers come to the restaurant and all the food magazines hype about - the bread is just one part of the á la carte meal. The restaurant pays the bills for the baker and makes it possible for him to give our bread for free.

Directory can be seen as a market (ecosystem) around the bakery. Someone builds a fountain (add-on) in front of the bakery to complement the free meal with free water. Next door shop starts to give out butter (add-on) for free for people eating Vaadin bread. Baker adds also some extras to his shop - maybe starts giving out cheese (add-on) for free with the bread. (All these are free add-ons in the directory - their availability doesn’t make the bread less free or make it taste less good - quite the opposite).

Many people like to have a glass (or two) of red wine with their meal - it just goes great together with bread (and free cheese). Some of those people even sell the meal in their restaurant for their customers. This is great business as the bread is good and free, but they can charge money for the meal from customers who just enjoy the meal - not the cooking process. Because no-one has been selling red wine in the market, the restaurant owners have been brewing wine on their cellars themselves, which is hard business (as in-house development often is).

Baker starts to sell red wine in his bakery. This bottled “Vaadin Zinfandel” costs a bit per cook in the restaurant serving it, but they are still free to serve unlimited number of bottles for any number of customers they want to without any extra payment. Someone else - who likes white wine more - starts to sell his “Mediacept Sauvignon Blanc” in the bakery. The baker is happy to provide space in the shop (Directory) for those excellent bottles of Sauvignon blank.

Now the people who have been eating free meals still can do so. In fact the free meal is better than ever. But now in addition to only drinking water with their meals (or brewing their own wine) they have an
of buying red or white wine with their meals.

I hope no-one of those people who prefer 100% free meal get jealous to someone who likes wine with their meals and start to demand Mediacept and IT Mill to either stop making the wine or give it out for free. Everyone can still continue brewing their own wine or have an option to pay for someone else (Mediacept or IT Mill) for the hard work of brewing wine.

In brief - adding possibility (for IT Mill and for others) to sell add-ons for Vaadin only
adds number of choices
and does not take any choices away. IT Mill, and I hope many others, will continue to invest (as heavily as ever) to the development of free Vaadin Framework as well as in development of free add-ons.
In addition to that investment
IT Mill, and I hope many others, will invest in building complementing add-ons that give even more choice to people using Vaadin Framework to build their applications.

So is there something that puts IT Mill into somewhat different position in the ecosystem than others? Yes, IT Mill still reserves commercial use of “Vaadin” name to itself (it is the sign in the front of the bakery). In exchange IT Mill has contributed most of the Vaadin and continues to contribute in the future. I hope that everyone finds this fair.

Hi Joonas,
Thank you for your analogy, definitely it’s now more clear for me, It was important to have your opinion, and now I agree with whatever you have said.