I don’t need to deal with javascript as a whole (including npm installation, compiling etc)
I am a backend oriented Java developer, go to ui developement only occasionally. I simply didn’t wanted, and don’t plan to ever go into the long, and frequently changing javascript framework learning curve. For quite a few actual working hours using the knowledge.
The move from Vaadin to abandon GWT and rebuild the platform on Polymer makes me feel, that I most likely never go beyond Vaadin 8, and probably go back to use GWT instead for ui purposes, without the otherwise very useful additions that Vaadin made.
So I would be very happy if there would be even a community supported Vaadin 9 project - continue the development of Vaadin framework with GWT, where the company turned other ways. I think I am not alone with this wish.
I just want to note that we are still supporting Vaadin 8, at least until February 2022. In practice that means doing frequent bug fixes and small improvements.
For example if there is going to be GWT 2.9 release before 2022 which adds support e.g. to JDK 11, we will evaluate if that can be used in Vaadin 8, and if yes we will include it in framework. It will most likely not make it Vaadin 9, but just Vaadin 8.X.
Note also, that using Vaadin 10+ does not mandate using Javascript. The big difference between Vaadin 8 and the new framework is that in case you want to implement custom components with client side, yes then some Javascript is needed. In Vaadin 8 you can implement the client side code also in Java due GWT (and also with Javascript if you so wish). My understanding is that most of the developers do not implement such components themselves, so it does not make much difference for them as long the availability of the add-ons in Directory is sufficient.
But just hypothetically speaking what would be your idea of “Vaadin 9”, what differences it would have compared to Vaadin 8? Is there some significant missing API, components, …
I don’t have a wishlist of Vaadin 9 features, but I suppose as development goes forward on the flow branch, there will be a number of features left out from Vaadin 8.
The concern is about if I would start a new longer term project, which can be estimated as a 3-5 years lifetime, then I won’t have support for Vaadin 8 after 2022 which is 3 years.
I did it for a small customer, and the plan is also about having more such smaller businesses, so getting enterprise support is out of scope.
Regarding ui components, once I needed to either create a better working html5 camera component, and also use extension to modify the upload component, in order to direct it to the camera application on a mobile phone. That required a little javascript programming, without any framework involved.
Interesting how you perceive Vaadin 10+. We certainly did not base Vaadin on Polymer, it’s just a helper library to build components but then again there are people who think Vaadin is a GWT extension which is not true either.
What we aim to do with Vaadin going forward is to offer value in shielding Vaadin users from the constantly changing JS world yet offer select capabilities that we’ll mature to new LTS release every couple of years.
Doing Vaadin 9 might seem tempting at first. It would continue full backwards compatibility with 8 series and we could see what features to add to it. It would be a safe bet for users like you. But Tatu asked an important question - in what way it would be different from 8.x releases? Do you just need extended support?
I used GWT previously, for mostly the same reason: java → javascript compiling, and shielding from browser differences. The most annoying thing with javascript development at that time (2009) was various browser’s different behaviour.
(Now, the frequent js framework waves add to this)
So that, when I discovered Vaadin, this was a ‘better, easier GWT’ for me.
The main concern is that with the turn of Vaadin to other js framework(s), there will be at best a temporal support for the Vaadin 8.x series. Having a plan for Vaadin 9, at least gives a message for longer term support that way.
I also understand that the lack of development on GWT makes a stressed situation for everyone who built an ecosystem over that. That’s why my idea was, that this could continue as a community effort.
they also have 2 major versions (1.0, 2.0) maintained over years, I suppose for the same reason: the 1.0 branch was too popular, users didn’t wanted to migrate
I do see your point with that. I think we have pretty much the same situation with V8 and V10+. It’s really up to you to decide which is better suitable for you.
I think V10+ has features that are quite compelling and it offers capabilities previously out of reach for server side Java developers. It’s not the same as V8 just as the version numbering should have indicated when we skipped V9. Vaadin will continue to evolve and we will keep our promise to offer innovative new capabilities of the web as they come available. I personally think that developing web apps without Vaadin is darn complicated business and there is plenty of room to innovate. Purists have plenty of choice in the field (of web dev). We are interested in offering DX that is truly better than what is out there today. We have now a modern base to build on and we’ll iterate fast.
Yes, 2.9 was released already a while ago. Java 9 - 11 support is tempting but this change: “Remove GWT version check.”, it is not drop in compatible in Vaadin.