so I’m doing polling instead.

which might be what’s killing my animation, I have no clue.

most animations while waiting are not animations like coded out in, in this case javafx, but are animated gifs

Yea i should do that

When the background requests are running, it jitters

you could lower the priority of the background thread

I'll try that

just went on a site and made a gif

easier than dealing with threads 😛

@pike I haven’t used nextgen but the previous version of openshift wildfly cartridge supported websockets

That’s a redhat service

@edvin This reproduces the runAsync problem from yesterday with the latest 1.7.13-SNAPSHOT. Running on my Mac, sometimes the “success message” does not show up. Only the first two println messages show. Run this 2 or three times. I slipped in an extra runAsync because I think the gap between what’s being reported in the field and what we’re working on in development is because of the small startup cost in the cached Thread executor. What I think is going on is that when the runAsync is primed by an earlier call, it can execute very quickly. Sometimes, this is before follow-up apply{}s containing the success block. There’s no synchronization barrier forcing the runAsync Thread to wait before finishing and is SUCCEEDED before we get to the apply.

I’ll try this on some other platforms today.

I was able to reproduce this just now — inconsistently — without the idomatic syntax.

@carlw uploaded a file: Extra Stuff Stripped Away from runAsync

does anyone out there have an example of binding a multiselect listview to a SimpleListProperty?

@edvin IMO, we’ll need to get the succeeded{} registered before the Thread is kicked off in order for this to work 100%.

@edvin commented on @carlw’s file https://kotlinlang.slack.com/files/U25U0KPFT/F8DC7SZU0/success_not_called_on_mac.java: Try with 1.7.14-SNAPSHOT 🙂

whoops…got the latest from my branch only 😊

@edvin Looks good with 1.7.14. I ran it a few different times on my Mac and didn’t see the bug in contrast with earlier (good A/B test). I’ll try the other platforms today in case the executor works differently.

@carlw Sweet, thank you! I feel pretty good about the fix, and I don’t think it’s worth optimizing the fact that we attach the listener in some cases where we know it won’t be run. This has the minimal amount of logic.

i feel good too, i'll hit the other platforms i have when i get a chance to see if the thread finishes differently. it looks like on the Mac, the state of the task is READY when your fix is hit and the setOnSuccess is called prior to it transitioning to SUCCESS

I'm SO CLOSE but SO FAR

@amanda.hinchman-dominguez uploaded a file:

I need to serialize the object it, but I think I have a few blockers:

1) datagrid is using a paginator list dictated by an observable list

2) I'm not sure if I can serialize the Tiles class itself, since TilesFX is simply a library I'm using and I can't actually extend the class

If I'm correct, an observable list is not serializable, might have to use an arraylist instead

I saw something promising here to create a wrapper class for the instance - thoughts? https://stackoverflow.com/questions/5256308/serialization-of-java-object-created-from-external-api

I casted my data class DragTile to be Serializable by the way