Call cb if provided to send even if callOnError was true

[anandthakker]

If I've created a send with createSend('blah', true) -- so that errors go to the store's onError handler -- I might also want to use send asynchronously. @yoshuawuyts what do you think about including an if (cb) { cb(err) } at the end of onErrorCallback (L131 on master)?

[yoshuawuyts]

@anandthakker could you write a quick patch for this? I think it'll make more sense once I see it - Main use case for passing true to createSend() was so reducers and views don't like need to call done() all the time.

Perhaps a possible solution would be to call createSend() twice? - but I'm keen to see what you mean. Thanks!

[anandthakker]

@yoshuawuyts yep, see ^. My use case here is that I want to use onError to set up an app-wide default error handler, even in cases where I might be using send asynchronously

[yoshuawuyts]

Ah, but so at the top level onError is always called; the first caller should generally come from a view (or technically onStateChange()) which gets the onError hook attached. If every caller propagates errors upwards using send(name, data, (err, res) => {}) and done(err, res) eventually the global onError() handler will be called - both with sync and async calls.

Does that make sense?

[yoshuawuyts]

@anandthakker is this still an issue? Any way I can assist?

[anandthakker]

Ah -- I've managed to move along okay without it, but it's sometimes
required the hack of passing a synchronous "send" as part of the payload to
an effect (since the effect's send is async). So, I guess that's just to
say, it's not a blocker for me, but I've still got a use case for it. Okay
with closing if this is not a direction you want to take it in

Why do I have effects that need a synchronous "send"? Because I'm actually
using a model to manage mounting / rendering / unmounting a bunch of
independent views, and so the "views:render" effect needs a sync send that
it can pass along to the view functions.

On Mon, Jul 18, 2016 at 8:27 AM Yoshua Wuyts [email protected]
wrote:

@anandthakker https://github.com/anandthakker is this still an issue?
Any way I can assist?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
#44 (comment),
or mute the thread
https://github.com/notifications/unsubscribe-auth/AEvmR2VKxqMNXoSC-HbUZ7HdixyFOn2lks5qW3EugaJpZM4JGAEb
.