@mikehearn @edvin I'm used to bringing my own services into a Java app (in fact i've build a career on it). However, if you look outside of the Java ecosystem, you'll see (at least w. Windows and UIKit) a different standard.

yes, microsoft and apple are willing to commit to providing cloud services long term.

ditto for google and android

@mikehearn oracle cloud services?

But there are so many standards. The service providers should (and do) provide API's to their services and client implementations for popular platforms. That's the way it should be IMO.

aren't targeted at consumers

i mean, oracle does not provide web maps, probably nor should they. they would not be as good as what google can do.

and indeed, apple's APIs insisting on using apple's services, microsoft's insisting on using bing, etc, can be annoying for end users even if it's helpful for devs

And the OS support is for their own platform, not for any standard, or even any other platforms.

i guess...but javafx gives you canvas ...microsoft and apple give you a maps control, etc

i'm looking for a broader desktop experience than just a double-click to a Stage

That's because the provide those services. And they only provide controls/apis for their own services.

yes, maps particularly aren't going to be a strength of the java platform anytime soon (maybe openstreetmap has a widget)

used to work on Google Maps

brb

@carlw broader experience should come from 'throwaway' projects on top of a solid javafx platform IMHO

@ron i want more components and features than what was developed for JavaFX 2.0

ControlsFX?

as an incubator

Yes

it looks like oracle jet has some of these controls so maybe there's some synergy there http://www.oracle.com/webfolder/technetwork/jet/jetCookbook.html?component=thematicMap&demo=default

My point wasn't to focus specifically on maps (my complaint was more about the System Tray) but it looks like there are standards like GeoJSON that might be blended into controls to make adoption more compelling

If any of you guys have any TornadoFX code that uses FXML, I'd be grateful if you can try the latest commit. I added some complexity even to the simplest cases, but the benefits are worth it, provided I got it right of course 🙂

@edvin If i create a few FXMLs for the test cases, is there a particular place you want them like a Kotlin package?

@carlw I just put the existing ones in

src/test/resources/tornadofx/tests

. There are only three there now, so no problem adding more.

@edvin I was testing the nested fxml and wrote a pair of test cases. Look for the PR. There are 3 fx:include scenarios to consider. 1. Controller per FXML. I use this all the time to treat my Java Controller as a code-behind. This means the fx:included files get their own Java class with -at-FXML annotations for fields and methods. 2. Top-level Controller (includes TornadoFX View). The topmost FXML gets a Controller or View and those top-level fields are mapped with fx:id. Controls in the fx:include are accessed by the scene graph. I recall being able to -at-FXML everything, even included fields, but I couldn't get this to work in a Java-only (no TornadoFX) test. 3. Outside Scene Graph. The fx:include is in a fx:define and is not part of the Scene Graph. A Controller class is needed because the define is disassociated from the parent FXML. This seems suited to standalone windows like dialogs. I've never used this in 4 years -- mainly because it might not be possible to do this solely in Scene Builder.

@carlw I'm working full-time on maps for JavaFX.

For lightweight mapping the web and JS is definitely the way to go and moves in a fast pace. Mapbox, OpenLayers and D3.js offer a lot. And then of course there are the Google Maps and Open Street Map APIs. Java has more to offer for professional mapping with JTS and GeoTools. Web maps were a long time only pre-rendered image (pyramids) and only in recent years they offer vector-based web maps.

@edvin do you mind if I add a parameter to fxlauncher to ignore ssl certificate errors?