A question about spec and re-frame. I think that it is quite useful to test that my subscriptions and handlers are conforming to a spec. Would it be feasible to add a meta-data key to subscriptions and handlers that specified input, and for subscriptions, output specs? Then one could 1 Instrument this in development 2 Run tests on each subscription in a test suite 3 Do generative testing (generating handler vectors) I’ll add an issue with this if it would be worth considering

@vikeri: or maybe extend re-frame with something like fdef, but for subscriptions/handlers?

@jstaffans Yes, that is definitely also an option

Has anyone come up with an error-capture/reporting interceptor, or how we’re supposed to handle that in 0.8? I see https://github.com/Day8/re-frame/issues/231 listed, but no responses so far.

e.g. the pattern I want is “try X; catch E; report E to somewhere else”, but that becomes harder now that interceptors do the actual invoking of the handler function somewhere else, as opposed to the previous user-land chain-of-handler-invocations, where try/catch was easy

https://youtu.be/DSjbTC-hvqQ?t=27m43s

I didn't see this, when I started re-frisk, but definitely imagined exactly the same functionality

hey party people!

im enjoying 0.8.0 but am a bit confused RE this: https://github.com/Day8/re-frame/blob/9a9ec3827b91513b741b4aeb2fe0781cc02d80c8/docs/Interceptors.md#the-builtin-interceptors

where are these built-in interceptors located?

i can't find trim-event anywhere in the re-frame 0.8.0 source

NVM. I didn't read. it's trim-v 😛

Unrelated: any way to get rid of this warning? https://puu.sh/rpDEg/8ad233c0b1.png

I know it's react complaining

JW if there's a workaround

why do you have that attribute on the input tag?

I'm just working off the examples

they have it structured like this

https://puu.sh/rpDKH/c89a1bb093.png

well, its hard to give any more help without any more reasoning behind why that might be there. I would just remove it and use a standard form :onSubmit, or maybe they wanted a button :onClick? hard to say.

https://puu.sh/rpDNc/2b103d5477.png

ah, [input {:on-save}] is not [:input {:on-save}]

input (not keyword) is a custom element defined somewhere else that accepts :on-save as a prop

whereas :input is a reagent keyword representing a standard html input tag

ahh

I see now

thanks @lwhorton

Hi, is re-frame 0.7.0 compatible with reagent 0.6.0 ?

I'm having trouble debugging one of my event handlers; it's logging out Handling re-frame event: ["asdf"]

shouldn't it be giving me.... Handling re-frame event: [:event-name "asdf"]

its hitting my correct handler but the db isn't updating is all

(reg-event-db
 :username-entered
 [trim-v debug]
 (fn [db [username]]
   (.log js/console "---> USERNAME ENTERED" username)
   (assoc db :username username)))

@samueldev trim-v quite literally removes the 1st element of the event vector. And it is running "before" debug.

You could always make the interceptors vector be [debug trim-v]

(ie. reversed)

That way debug will be logging the event before trim-v alters it

Gotcha, that fixed that

But I'm still getting no app db changes triggered by... even though my handler is pretty plain:

(reg-event-db
 :username-entered
 [debug trim-v]
 (fn [db [username]]
   (.log js/console "USERNAME SET" username)
   (assoc db :username username)))

@lwhorton No good solution currently other than to hook https://developer.mozilla.org/en/docs/Web/API/GlobalEventHandlers/onerror

@mikethompson able to point me in the direction of any debugging steps RE above?

@samueldev it appears your old db is testing = to your old db

Not sure why

@lwhorton a small update: https://github.com/Day8/re-frame/issues/231#issuecomment-249979258

@johanatan yes, re-frame 0.7.0 is compatible with reagent 0.6.0

@andre I'd very much like to add an intrumenting interface to re-frame. One which would allow a piece of code to be told when certain events in re-frame happen. For example "starting to handle event [:a :b]". Running subscription [:get-user 42] Then, this code could build up a data structure which captures this instrumentation data. Which itself could then be shown via something like data-frisk.

Obviously the code which is capturing the instrumentation data doesn't have to simply record it (otherwise good old console is good enough) but instead can build up a custom data structure.

@mikethompson: okay, thanks. Do you know if both will work with the latest version of React?

i.e., if I use :exclusions to prevent them bringing in React and then manually bring in the latest version myself

re-frame doesn't depend directly on React, so the question is more if Reagent 0.6.0 will work with the latest React, and the answer to that is yes

@mikethompson awesome, thx!

wonder if anyone got issues like: dispatching will invoke handler twice in dev, but not in prod cljsbuild?

tried to do git bisect to find out what change caused it. no luck