Hi,

sorry for being late tot he party but I agree with whats been said and would 
opt for 1 or 3.
We use Vaadin quite a lot and since Vaadin 10 or "flow" it does no longer 
require GWT so theres no necessity to change anything in the build (and it 
integrates nicely with Spring).
But Option 1 is surely also fine.

Julian

Am 15.10.19, 17:02 schrieb "Tim Kläsener" <[email protected]>:

    Thanks for your thoughts and points Chris and Kai!
    
    I think I'll take your suggestions and start/continue with the plain
    HTML/JS approach of option 1 as I already started and as the assets
    are served from Spring Boot itself so integration into the build is
    quite easy. If this gains traction I have a couple of blueprints to
    migrate over to Vue or Angular (or even JSF, if necessary) ;-)
    
    Cheers,
    
    Tim
    
    
    On Tue, 15 Oct 2019 at 15:01, Kai Wähner <[email protected]> wrote:
    >
    > I was just smiling a little bit while reading this because I remembered 
my presentations almost 10 years ago about such discussions...
    >
    > Just for reference (you don't need to agree with my categorization), and 
today you think much more about "Vue/Angular/React/JS-Framework of the week". I 
discussed the following ~10 years ago:
    >
    >
    > 
http://www.kai-waehner.de/blog/2010/12/30/categorization-of-web-frameworks-in-the-java-environment/
    > 
https://www.slideshare.net/KaiWaehner/categorization-of-java-web-frameworks-jsf-wicket-tapestry-gwt-zk-flex-javafx-grails-lift-spring-mvc-roo
    >
    > From my experience, I would clearly vote for your option 1) =>
    > Simple handwritten html+javascript (Jquery?) served by the Spring Boot 
backend (easy to start with but might get a bit messy over time)
    >
    > If the UI gets a lot of traction and evolves over time, some refactoring 
(and the inclusion of a web framework) is probably reasonable. But I would 
definitely start with a "quick and dirty" keep it simple solution which just 
works. A web framework makes it much more complex from the beginning. I had a 
project with GWT... GWT was awesome, but also the biggest pain in the project 
becasue most people of the team did not understand the framework in detail... 
And the UI is not they main part of PLC4X.
    >
    > Start simple, if it works well and gets traction, plan a good web 
architecture without pressure in parallel and release it in another major 
release of this Apache project. I guess most other Apache projects started 
similarly (even today most UIs are not really nice).
    >
    > Best regards,
    > Kai Waehner
    >
    >
    > Just my two cents from my experience.
    >
    >
    >
    >
    > On Tue, Oct 15, 2019 at 1:53 PM Tim Kläsener 
<[email protected]> wrote:
    >>
    >> Hi everyone again,
    >>
    >> We're currently looking into creating a frontend for interacting with
    >> the PLC Simulator, i.e. having an administration interface for
    >> changing properties/input/output and, if time allows, some fancier
    >> visual representations of the devices the PLCs control.
    >>
    >> As for the REST service, I've so far developed it with Spring Boot
    >> which should be okay from a licensing perspective (let me know if you
    >> have any concerns), now I'm looking into different options for
    >> developing the frontend:
    >>
    >> 1. Simple handwritten html+javascript (Jquery?) served by the Spring
    >> Boot backend (easy to start with but might get a bit messy over time)
    >> 2. Separate single page application using
    >> Vue/Angular/React/JS-Framework of the week (better separation of
    >> concerns but requires JS-toolset)
    >> 3. Java-based web UI framework like Vaadin or Royale
    >>
    >> Any other suggestions or reasons to choose one option over the others?
    >>
    >> Cheers,
    >>
    >> Tim
    

Reply via email to