Fork me on GitHub
#re-frame
<
2020-05-05
>
mikethompson03:05:29

@kenny can you add your thoughts to the issue itself

mikethompson03:05:52

As isaac indicated when closing the issue, we closed it primarily because we didn't hear anyone saying it was a problem any longer. We have a plan for doing but only if we need to.

kenny03:05:23

I was curious about the potential implementation of the point I quoted 🙂

mikethompson03:05:29

Maybe this ... The proposal is that an event will now be queued as map of values. Each map in the queue could have a string :source field which EITHER indicates the source code coordinates or it could say "dispatched as an effect when processing event [:id arg1]"

kenny03:05:18

And that would need to be manually attached to the event or some macro magic would do it?

mikethompson03:05:51

Because there would be no source code coordinates, no need for macro magic. You are just getting the event which triggered the :dispatch effect

mikethompson03:05:38

But these are just ideas at this point

mikethompson03:05:49

Anyway, if you believe it is important, then please just make the case on the issue. We'll review

p-himik04:05:52

Given that there are some people working with the queue manually, maybe it's possible to not switch from the vectors but just add some metadata to them instead?

mikethompson05:05:59

@p-himik Could you explain more about who/what might be working on the queue. I've through about it as an internal implementation detail ... but are you saying there are some with a dependency.

p-himik05:05:04

No, nothing concrete that I know of. But ~once a month someone asks about something advanced and then the solution is to go fiddle with the queue manually.

p-himik05:05:46

Maybe it doesn't justify any attempt at preserving having only vectors in the queue. But apart from that, the aforementioned :source field does sound like metadata to me.

mikethompson05:05:20

Hmm. Yes, it can be a fine line between what's data and what's metadata

mikethompson05:05:06

"Eye of the beholder" almost

4
p-himik05:05:06

There's also add-post-event-callback that accepts the queue itself in full.

p-himik05:05:43

Doesn't seem like re-frame.router/event-queue is used in any public GitHub repo. Searched for "re-frame" "router" "event-queue" - only re-frame sources themselves are there.

mikethompson05:05:14

Yeah, i just did a quick search too.

p-himik05:05:02

But add-post-event-callback is definitely used, at least by re-frisk.

mikethompson05:05:36

It is a very obscure piece of the API. But backwards compatibility might still be possible. Anyway, at this point, we're not making a change, unless we get some usecases in the (now closed) issue

p-himik05:05:02

Sure. A discussion doesn't hurt. :)

mikethompson03:05:54

@wvelezva wait, now I see. You are specifically after testing. The site I provided is perhaps not the best then.

mikethompson03:05:46

---------- Isaac and I are doing a three week long sprint on re-frame - starting yesterday. Isaac's time is being paid for by ClojuristTogether funding. My time is paid for by Day8. If you have any ideas / needs etc, now is the time to talk up (in the issues please)

mikethompson03:05:19

@p-himik universal interceptors, yes, on the list.

😄 12
❤️ 12
cljs 8
Setzer2207:05:39

I'm having trouble setting up a very simple example in my application. It just won't do anything :S

mikethompson12:05:52

If you want to keep up to date with the current re-frame sprint, please keep an eye on this channel and this issue https://github.com/day8/re-frame/issues/589

bmo 12
❤️ 24
👍 8
oliy12:05:48

i'm not familiar with the meaty issues, are they documented anywhere?

👆 4
victorb14:05:08

one could imagine that some of the ones here (ordered by # of comments) are the meatier issues: https://github.com/day8/re-frame/issues?q=is%3Aissue+is%3Aopen+sort%3Acomments-desc

mikethompson14:05:15

They are somewhat covered in these incomplete documents https://github.com/day8/re-frame/tree/master/docs/EPs Part of this process is to flesh this out.

victorb07:05:11

@mikethompson some questions about one of the EPs, if you don't mind. How serious is the 005-StateMachines proposal and why would it be beneficial to have that in core? There been a couple of experiments of third-party libraries, both for FSM in general, and also Statecharts, I'm guessing you've seen those in the past as well? Asking because I'm interested in trying to figure out how to leverage Statecharts better in re-frame applications, but unsure I understand the reasoning behind the "why in core" when it comes to that specific proposal

victorb07:05:40

or, my assumption that "EP" => "Proposed change to day8/re-frame" is faulty, and any EP could be implemented either in core or as separate library

David Pham17:05:42

Good luck @mikethompson for the 3 weeks sprint.

David Pham18:05:53

@superstructor Also good luck! Thanks a lot for your help!

David Pham18:05:20

I love your documentation for re-frame, your readme was extremely funny and I still remember I had a great moment readin git.

🙂 4