Eventually should allow configuration for space between requests

[LeoColman]

No description provided.

[matheusmessora]

Based on Awailitiy, this is called Pooling.
Also, should be great if we create a default configuration from Duration and Pooling.
At the moment, eventually calls a while(true) method.

[sksamuel]

Yeah we must improve eventually. I looked at awatility and it seems to be basically the same as our eventually.

I think what we need to support are different durations, different polling methods, and what awaitility calls conditional evaluation listeners. https://github.com/awaitility/awaitility/wiki/Usage#condition-evaluation-listener

Happy for anyone to pick up this or I can tackle it.

[matheusmessora]

I'll try to work on that

[LeoColman]

We're here to help if needed.

[matheusmessora]

The actual eventually receives all the data inside the method parameters. For a non-breaking change, we can add more parameters with default values.

My suggestion is to create a "new" eventually. Maybe eventual returning a factory with customizable methods. Similar to awaitility.
This will improve the extensibility of the eventual function.
The polling interval by the default is already configured.

The result could be something like this:
eventual() { 1 == 1 }

eventual().during(ofSeconds(2)) { 1 == 1}

eventual().during(2).ignoreExceptions() { 1 == 1}

eventual().ignoreExceptions(IOException.class) { 1 == 1}

What do you guys think?

[LeoColman]

Perhaps we can use the eventually construct, but deprecating the current eventually in favor of the builder alternative, and remove it in the future.

But I believe that using a builder is too much Javalistic, and we should come up with a better DSL for it (I don't have any suggestions yet :P)

[matheusmessora]

Do you have a example of a good one?
PS: I came from Java. Don't judge me

[sksamuel]

eventually(polling = exponential(8), delay = 4.seconds, ..... and so on) { } Then you can do eventually { } If you're happy with the defaults. And it's 100% backwards compatible.

…

On Fri, 14 Jun 2019 at 12:57, Matheus Messora ***@***.***> wrote: Do you have a example of a good one? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#793>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAFVSGR3NGGDXHQJKJRTUDDP2PLYFANCNFSM4HQTN4WA> .

[LeoColman]

I liked your example, @sksamuel

[sksamuel]

I would like to see this in 3.4 so I can pick this up if it's stalled.