cool. night 🙂

Pushed 2.6.0-RC5 to Clojars. I ran through all of the book demos and found some minor problems. It turned out that a few of the demos had bugs that “made it through” and worked, but which didn’t continue to work in the new version. It is also interesting to note that the changes to component local state (we no longer wrap this, it is direct React) can cause minor differences that might affect you.

@tony.kay when i try to run lein repl with RC5 i get this exception

Caused by: java.lang.RuntimeException: No such var: fulcro.client.dom-common/interpret-classes, compiling:(fulcro/client/dom_server.clj:439:1)

it doesn’t happen with RC4

hm

open an issue I guess

you’re not seeing it?

might be something with my deps then

is there a way to see the diff between RC4 and RC5 in github?

maybe i can try to figure it out that way, it only happens if i require the dom-server anyway

i guess i can just look at the commits from Merge tag '1.6.0-RC4' into develop to the RC5 one

weird thing is i’m relying on fulcro.client.dom-common/interpret-classes in the client and its working fine, but it’s not defined in the server side

I’m working on other things…so no time to look right now

SSR has not been tested much in 2.6, and is likely to have issues still

React 16 is doing SSR differently, and we need to patch that (new hydrate call)

and I think we get to drop checksums

no worries, i made a new project from template, added my deps, and required dom-server in there, it worked fine

so most likely something in my code

:man-facepalming:

@tony.kay lein clean was all i needed to do

it’s working fine

well that was a “fun” waste of time, maybe i should just get a tattoo that says lein clean?

😄

this is one of the advantages of boot’s immutable file system, there’s never a need to boot clean

I think I'm having another issue with set-state! being not immediate, I'm using setTimeout in a handler so I guess that might end up overwriting set-state! somehow. From this point of the react state docs I understand I could pass an updater function: https://reactjs.org/docs/faq-state.html#how-do-i-update-state-with-values-that-depend-on-the-current-state But that doesn't work with prim/set-state!. Ended up fixing my issue like this:

(.setState this-of-guide  
  (fn [s]  #js {"fulcro$state" (dissoc (aget s "fulcro$state") :started-eval-step)}))

@bbss So, set-state! uses the updater function…perhaps there is a bug in it?

https://github.com/fulcrologic/fulcro/blob/develop/src/main/fulcro/client/primitives.cljc#L2417

oh, I see: you want to remove something from state

Yes, might be. It isn't obvious to me why it wasn't working.

the behavior of set-state! uses, as you can see, the exact pattern, but it assumes you’re passing data to merge with existing state (as setState does)

update-state! works more like swap!

https://github.com/fulcrologic/fulcro/blob/develop/src/main/fulcro/client/primitives.cljc#L2432

oh wait…`update-state!` does look like it has a bug…it should not be using get-state

I’ll open an issue on that

I should have asked you for what you were doing/expecting

@tony.kay Thanks, I'll have a look at update-state! tomorrow and try to explain what I'm doing and maybe figure out why it didn't work.