Clojurians
#cljs-dev
<
2017-05-26
>

This page is not created by, affiliated with, or supported by Slack Technologies, Inc.

rauh08:05:31

Should MetaFn implement ICloneable?

dnolen12:05:49

@rauh I don’t see any good reason for that

rauh12:05:00

@dnolen (specify some-meta-fn ...) fails without it. Not a common use case TBH, so I'd agree that it's not a good reason.

dnolen12:05:55

I don’t see any reason why specify should work on actual fns

rauh13:05:32

@dnolen Could you elaborate on the MetaFn/DCE comment you just wrote? I'm confused.

dnolen13:05:59

need more information

dnolen13:05:13

adding protocol indirection is just degrading performance for saving how many bytes?

dnolen13:05:31

if it’s 1K in typical projects, this isn’t important

rauh13:05:42

But where would we add protocol indirection? If we substitute with-meta for -with-meta?

rauh13:05:18

I mean, the protocol gets called in with-meta anyways, so we save a function call, no?

dnolen14:05:13

oh right, brain lapse, sorry I was thinking those with-meta were method calls but they aren’t

thheller17:05:05

@dnolen are you doing a new build/release today?

dnolen17:05:30

wasn’t planning on it, there’s nothing particular important that’s come through

thheller17:05:28

I’d like to get my hands on a release with CLJS-2040

thheller17:05:30

but no rush

anmonteiro17:05:08

@dnolen I’m just preparing a patch with the new changes to spec that landed today

anmonteiro17:05:25

not sure if it warrants “important” or not though :slightly_smiling_face:

dnolen17:05:10

ok, could do for spec changes

anmonteiro17:05:27

btw I also upgraded Closure to the May release as part of https://dev.clojure.org/jira/browse/CLJS-2027

anmonteiro17:05:18

@dnolen I think we should also consider upgrading Closure Library again. Someone mentioned they needed something that had been released in a newer version a couple days ago

anmonteiro17:05:37

(not for this release)

dnolen17:05:23

yeah ok, I can trigger a build of Closure Library

dnolen18:05:25

would be nice to get some verification that this patch doesn’t affect source mapping https://dev.clojure.org/jira/browse/CLJS-2051?focusedCommentId=45737#comment-45737

frank18:05:20

would also love to have a release with https://dev.clojure.org/jira/browse/CLJS-2021 (I'm on the list this time)

dnolen18:05:00

@frank patch won’t apply, did you follow the patch creation instructions?

dnolen18:05:12

Patch format detection failed. is the message from git am

frank18:05:21

ahh oops didn't see those. I'll fix it now

frank18:05:12

submitted a proper patch this time

dnolen18:05:27

@frank thanks, applied

dnolen20:05:25

I just cut org.clojure/google-closure-library and org.clojure/google-closure-library-third-party version 0.0-20170519-fa0499ef, might want to take that for a spin

anmonteiro20:05:15

trying it now

anmonteiro20:05:32

@dnolen small typo above (messes with copy-pasting) org.closure->`org.clojure`

dnolen20:05:29

@anmonteiro thanks fixed

anmonteiro20:05:47

sorry, also google-closure-third-party -> google-closure-library-third-party

anmonteiro20:05:02

all tests pass downstream in Lumo, but we don’t use the Closure Library a great deal

dnolen20:05:12

@anmonteiro still that’s good to know

anmonteiro20:05:21

same thing in Om

dnolen21:05:31

unless someone has an objection I’m going to cut a release of ClojureScript now

anmonteiro21:05:20

@dnolen did you get to CLJS-2027?

dnolen21:05:52

oops didn’t

dnolen21:05:54

doing that now

dnolen21:05:17

@anmonteiro actually needs a rebase

dnolen21:05:30

am -3 wouldn’t work

anmonteiro21:05:30

doing it now

dnolen21:05:04

@anmonteiro didn’t work :slightly_smiling_face:

dnolen21:05:11

there’s a lot of stuff in that patch from previous commit

anmonteiro21:05:26

oh I messed up

anmonteiro21:05:36

@dnolen sorry about that, fixed now

anmonteiro21:05:10

also, please hold up on the release, @mfikes found a slight problem with self-host

anmonteiro21:05:07

patch attached

dnolen21:05:05

@anmonteiro got it

mfikes21:05:43

The patch fixed it for JVM, but in self-host, I got

WARNING: Use of undeclared Var cljs.spec.alpha/any? at line 34 src/planck/core.cljs

mfikes21:05:44

Odd. I wonder if it is meant to refer to the core any? predicate…

mfikes21:05:26

Syntax quote is probably broken in self-host

mfikes21:05:51

I’ll file a separate self-host issue. This shouln’t block a JVM release.

anmonteiro21:05:28

I could also edit this ticket to read cljs.core/any?

dnolen21:05:51

@mfikes @anmonteiro let’s just fix this, just namespace that symbol right?

mfikes21:05:18

Yeah, was thinking that too. If you literally put in 'cljs.core/any? it might do it

dnolen21:05:02

@anmonteiro applied the ES7 thing

dnolen21:05:19

er 2017 or whatever it’s called :slightly_smiling_face:

dnolen21:05:53

@mfikes which test are you running for self-host the self-parity one?

dnolen21:05:00

I’m here and can fix the issue now

mfikes21:05:10

I just noticed this downstream in Planck unfortunately

mfikes21:05:40

If you are thinking about fixing it, I can easily confirm

dnolen21:05:31

@mfikes done

mfikes21:05:33

Cool. Testing now.

mfikes21:05:59

@dnolen That change looks good. I'm on to other errors in Planck now, which I think are cut-n-dry things where I need to just update for the latest Google Closure namespaces.

mfikes21:05:05

I'll sort out a minimal repro for the fact that syntax-quote is somehow broken for this case in self-host, and file a JIRA, etc.

ewen21:05:21

@dnolen I forgot a call to "with-meta" in core.cljc for https://dev.clojure.org/jira/browse/CLJS-1884, should I reopen the issue ?

dnolen21:05:43

@ewen just point it out to me now and I will fix it

ewen21:05:15

thx @rauh for pointing it out

dnolen21:05:07

cool, changed here & testing now

dnolen21:05:44

cutting a release now

ewen21:05:56

thx :slightly_smiling_face: