Fork me on GitHub

@hlship Mainly just that Paul (the primary steward until now) has to devote his time to personal life for the foreseeable. I'm the maintainer at the moment with a pretty full plate.


But I sincerely appreciate the prodding. No sarcasm!


I'd love to hear from @channel what they need from Pedestal. IMO, the foundation is strong so I don't see a lot of pending work in the core. So post your wishlist and we can start to build a roadmap together.


Yep. I've just had zero time for any open source work the last few weeks.


@mtnygard one of the things I’d like to help with is fleshing out the guides more. It can be frustrating when learning pedestal to click on a guide link only to see a placeholder template. I’ve thought about contributing some but notice that PRs seem to stagnate, so haven’t put in the time yet.


I would love help with the guides.


Great! I’ll see where I can help and will work on some PRs over the next few weeks. Would you be open to a PR that removes any link to a guide that does not yet exist (to remove the tease) for the time being?


Well, I did put in the placeholders to encourage contributions, not just to tease people.


But if your experience is that they're discouraging, then go ahead and send that PR.


What if a general call to action was put at the top of the guide list?


I like that


The list would then act also like a todo list for contributors


I would like to see the interceptor model extracted into a specification, like Ring: no pedestal-based keys, pluggable async and support for ClojureScript.


@ikitommi Are you more interested in using interceptors in front-end apps or in back-end CLJS? Just curious.


both. re-frame has interceptors, but bit different, (`:before` and :after and with bit different semantics) and would be great for Clojure/Script story to be able to run ~same web apps on both server targets.


There are at least two ports of pedestal interceptors, just bit different.


we have some libs that suppot both ring and “Pedestal-style intereptors”.


I found that, when implementing GraphQL subscriptions for Lacinia, using Pedestal web sockets, there existing API was quite opinionated and I had to do a kind of "end run" around the preconceptions:


It feels like there's a simpler API buried in there, and the current API should be a convenience layer on top of that.