This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2018-12-04
Channels
- # adventofcode (100)
- # announcements (7)
- # architecture (1)
- # aws (14)
- # beginners (209)
- # calva (30)
- # cider (5)
- # cljdoc (2)
- # cljs-dev (37)
- # cljsrn (2)
- # clojure (133)
- # clojure-dev (20)
- # clojure-finland (1)
- # clojure-italy (10)
- # clojure-nl (19)
- # clojure-spec (56)
- # clojure-uk (49)
- # clojurescript (57)
- # clojurex (8)
- # core-async (2)
- # core-logic (1)
- # cursive (38)
- # data-science (19)
- # datomic (28)
- # devcards (3)
- # duct (8)
- # emacs (28)
- # figwheel (1)
- # figwheel-main (31)
- # fulcro (2)
- # jobs (1)
- # kaocha (1)
- # klipse (2)
- # mount (6)
- # nrepl (43)
- # off-topic (20)
- # pathom (3)
- # pedestal (1)
- # re-frame (15)
- # ring-swagger (1)
- # shadow-cljs (47)
- # spacemacs (19)
- # sql (20)
- # tools-deps (58)
- # unrepl (13)
- # vim (5)
@dnolen any objections before I attempt to fix this? https://dev.clojure.org/jira/browse/CLJS-3003
@thheller I'm a bit confused by the report since I thought we already did that, in fact your advanced snippet at the bottom does what you say no? Dispatch on arity and invoke the right methods?
@dnolen no. it just happens to delegate to cljs.core/get
in both cases. in the keyword case thats all the impl does so it looks like it is doing the right thing. if the actual impl is longer the full code is repeated.
demo.browser.Repeated.prototype.call = (function (self__){
var self__ = this;
var self____$1 = this;
var _ = self____$1;
console.log("repeated");
return self__.a;
});
demo.browser.Repeated.prototype.cljs$core$IFn$_invoke$arity$0 = (function (){
var self__ = this;
var _ = this;
console.log("repeated");
return self__.a;
});
@mfikes I can't figure out why the patch for https://dev.clojure.org/jira/browse/CLJS-3003 fails the self-host tests. maybe you have an idea? can't figure it out. the generated code seems fine for non self-host.
With that change, I see an error in test-cljs-2133
and a failure in test-cljs-2741
when running script/test-self-host
. Perhaps you've sorted that?
new patch doesn't use ->>
anymore. but the issue was that it always started with arity 0..n even if only n was actually declared
I put that new patch through CI / Canary. By the way, frequently updated patches are attached instead of replacing the originals.
I typically delete patches to avoid confusion which the correct one is but I can attach as instead if preferred
Cool. We may have clarfication on what to do via https://github.com/clojure/clojurescript-site/pull/266 Or, we can switch to using GitHub.
Looks like that last patch fails with https://gist.github.com/mfikes/814402fb3eeaec8a6a5500059c94061a
I got an arity error with the last one: https://gist.github.com/mfikes/36d17e466a5f188f325c22ff11109009
@dnolen my change in CLJS-3003 breaks https://dev.clojure.org/jira/browse/CLJS-2133. should this still actually be supported given that it officially isn't supported?
My vague memory on that one was that we reverted it so as to avoid lots of downstream failures by Reagent. Hopefully in the intervening year+ Reagent fixed it...
I guess we can treat this as a deprecation period and finally remove it? seems weird to keep it around forever. I can however adjust the patch if requested.
I wonder if David is thinking of releasing soon to match Clojure 1.10. If so, perhaps CLJS-3003 could be applied soon after.
@mfikes do you need to trigger your CI stuff manually? If so please do for https://dev.clojure.org/jira/browse/CLJS-3002