anyone used open layers and externs?

@grav open layers is a closure library, you should be able to use it without externs

@martinklepsch: well it doesn’t work in advanced compilation.

@grav: how do you use it?

https://twitter.com/conanistweeting/status/661502363241435136

I refer to it as eg js/ol.style.Circle., and it changed my code to something along the lines of new ()

@grav: how did you include it in your project?

I have no clojure-specific setting for it, eg. I just load it from ol’s cdn from my index.html

<script src="http://openlayers.org/en/v3.10.0/build/ol.js" type="text/javascript"></script>

@grav: this will help: https://github.com/cljsjs/packages/tree/master/openlayers

@grav: also search the issues in this repo for additional discussion

@martinklepsch: I tried that, but I may have used in incorrectly. Will try again, and report back

@grav there are people using this in production I think. IIRC OL is pretty big so you really want dead code elimination etc.

For trying out lein-figwheel 0.5.0-SNAPSHOT, could I do lein new figwheel foo then just change the plugin version? Or do I need to do something prior to generating the project?

changing the plugin version should be good enough, just make sure that if any of the figwheel options have changed, that you change them in project.clj as well.

thanks

hmm packaging openlayers via cljsjs seems like a terrible idea, considering it is already closure compatible js

@thheller: it’s packaged as a :lib instead of :foreign-lib

this way you can pull it in via maven in a form that the clojurescript compiler understands without any extra effort on the side of the user

ah though it was using the advanced compiled version that openlayers ships

instead of the full source

no, it’s full closure compatible source

which makes it one of the nicest to use libs in cljsjs I think

yeah closure-js is much better integrated than common-js&co

wow openlayers looks very cool

hey guys, newbie question from a first time submission (http://dev.clojure.org/jira/browse/CLJS-1411)

the wiki says "Edit the ticket's Patch field to indicate either "Code" or "Code and Test". This makes it easier for others to find patches in need of review."

I'm not sure what this means?

is there something else I need to do or should I just wait ..?

@anmonteiro: there’s a #C07UQ678E channel

did not know that

thanks

is there a way to use your own macro with cljs-bootstrap-compiled code? say

(defmacro unless [pred a b]
  `(if (not ~pred) ~a ~b))

(unless false (println "Will print") (println "Will not print"))

the code above compiles into the following:

cljs.user.unless = (function cljs$user$unless(_AMPERSAND_form,_AMPERSAND_env,pred,a,b){
return cljs.core.sequence.call(null,cljs.core.concat.call(null,cljs.core._conj.call(null,cljs.core.List.EMPTY,new cljs.core.Symbol(null,"if","if",1181717262,null)),cljs.core._conj.call(null,cljs.core.List.EMPTY,cljs.core.sequence.call(null,cljs.core.concat.call(null,cljs.core._conj.call(null,cljs.core.List.EMPTY,new cljs.core.Symbol(null,"not","not",1044554643,null)),cljs.core._conj.call(null,cljs.core.List.EMPTY,pred)))),cljs.core._conj.call(null,cljs.core.List.EMPTY,a),cljs.core._conj.call(null,cljs.core.List.EMPTY,b)));
});

cljs.user.unless.cljs$lang$macro = true;
cljs.user.unless.call(null,false,cljs.core.println.call(null,"Will print"),cljs.core.println.call(null,"Will not print"));

@settinghead: the rules about macros are not different between ClojureScript JVM and JS

it must be in a macro file

I've been pondering, and I think the self-host macro rule ends up being: “To call a macro, it must be defined in a namespace required as a macro namespace.”

@richiardiandrea: Hey I fixed the repl-api and I just re-deployed the SNAPSHOT. Make sure you do a lein clean before trying it.

ok, trying!

@bhauman: works now!

yeah it was really broken

easy to tell it's not a code path I use often

thanks for the heads up on it

np any time

I get the same issue still

is there a reason lein clean could not work as expected?

I didn’t see any downloads happen after the lein clean and then running lein figwheel

@jeremyraines: lein clean only works if you specify :clean-targets

what do I add there to have it clean dependencies?

@jeremyraines: what platform are you on?

osx

you should have seen the new SNAPSHOTs download

you may have to delete the figwheel snapshots manually from your ~/.m2/repository

ok

and then start figwheel

I was advising you to run lein clean to get rid of Clojure cached classes

@jeremyraines: it’s just a vector of file paths

ok, wasn’t sure if it was advisable to have it reach outside the project

removing it from ~/.m2 made the downloads happen, but I get this error

Exception in thread "main" java.lang.RuntimeException: No such var: figwheel-sidecar.repl-api/start-figwheel-from-lein

yeah you didn't get the new plugin

make sure you deleted lein-figwheel as well

after that, lein clean lein-figwheel, same error

sorry, forgot to delete the others from m2

works now

is there a way to connect this method with the system stuff yet, for example to get CSS watching?

@jeremyraines: CSS watching works you have to configure it

:figwheel-options :css-dirs

awesome

oh it's been changed to figwheel-options

@bhauman: is it figwheel-options both inside and a cljsbuild and inside lein project?

Is there a more pretty way to map a list of things (e.g. numbers) to a list of maps than (map #(-> {:id %}) [1 2 3 4]), i.e. something more elegant than ->?

jannis, I’d tend to do what you’re doing more explicitly, simply using (map #(hash-map :id %) [1 2 3 4])

@jannis lots of way for that specific case

(whoops)

(map (partial hash-map :id) [1 2 3 4])

depends on your idea of elegance

(zipmap (repeat :id) [1 2 3 4])

yeah

dnolen, that doesn’t work

oh right sorry

not one map, N maps

but, similarly you could do (map hash-map (repeat :id) [1 2 3 4])

not sure that’s an improvement or not…haha

code golf! 😛

heh

Haha, okay

I think Jannis’s original was the shortest actually!

at least, I can’t figure out how to make it more concise

I guess #(-> [:id %]) is fairly obvious and you can see the structure of the resulting maps right in front of your eyes.

I would recommend whichever formulation “future you” will grok the most readily

-> is noise in this case

yeah, my only argument against the threading macro is that it’s not really any more clear here

yeah

(map #(hash-map :id %) [1 2 3 4])

christianromney, yeah.

(re: future you comment)

provided you’re not on a team with some collective idioms already established

@christianromney: Yep, good point. @dnolen: That's nice, too.

(map (partial hash-map :id) [1 2 3 4] But maybe a little bit more obscure.

that’s a case in point. my brain naturally gravitates to partial in these cases even though #() is shorter

yeah, I find partial more explicit, but it requires knowledge of partial. Might be harder for beginners.

of course it depends on arg position

oh, yeah, that too

personally I never use partial and suspect it’s as verbose as it is to encourage people to just use fn literals

also if you’re a perf nut depending on the compiler there are implications for partial vs. inline fn literal

but agreed it’s most a readability / intention thing

when you’re using partial, it doesn’t really make it a partial like it haskell? its more or a hack?

“doesn’t really make it a partial"

what does this mean?

like, does partial in clojure have the same power as partial functions in haskell?

I don’t understand - partial functions don’t have anything to do with haskell

it’s just a functional programming concept

interesting

so partials in clojure works as any other partial function in other functional languages?

It's a partially applied function, no different from any other function

You can do with it whatever you can do with functions otherwise

haskell’s evaluation strategy is fundamentally different from Clojure’s, being lazy, so you can’t really compare the two tmtwd

interesting

it is analogous to a partially applied function in Haskell I suppose, but you don’t need a special function for it in Haskell

@tmtwd: I'm not sure if this answers your question, but 'partial' only uncurries the arguments it needs to. You could write code like in haskell but you'd have to put partial in front of every single function!

right

I’ll have to look into it some more

thomasa, I’m not sure I understand what you mean by “put partial in front of every single function”…are you talking about what you would have to do in Clojure to imitate Haskell’s behavior?

yes

I see

actually, sorry I don't mean every single function, i mean every fn with more than one argument

jannis: I would probably write that (for [i [1 2 3 4]] {:id i})

@lvh: Not bad either

amazing, http://jannis.github.io/om-next-kanban-demo/cards.html#!/

Anybody know when the next release of clojurescript is going to be released? I am seeing java.lang.NullPointerException at com.google.javascript.jscomp.JSModule.sortInputsByDeps(JSModule.java:263) It appears this is potentially resolved with CLJS-1469 :modules regression

shooting for this Friday

excellent, thanks

@dnolen: wouldn't be awesome to add a layer on top of figwheel with cards and a in-brower repl? Voila' we have a ClojureScript IDE 😄

@richiardiandrea: I already have a ClojureScript IDE it’s called Cursive 😉

but yes, something like that would be cool for it’s own reasons

very cool indeed

@richiardiandrea: :figwheel-options is not to be used inside the project.clj Its only for calling figwheel-sidecar functions that kick off figwheel