Re: clojurescript closure problem?
I stand corrected. I guess I'm *too* familiar with JS. I'll have to re-learn a few things. m On Thu, Sep 22, 2011 at 8:34 AM, Meikel Brandmeyer (kotarak) wrote: > Hi, > > Disclaimer: I haven't touched ClojureScript a single time up to now, I only > know the talk of Rich about it. Neither am I JavaScript developer. > > Too me this smells like a bug. ClojureScript is not JavaScript. It tries to > bring (where possible) Clojure semantics to the JavaScript VM. The first > version is perfectly valid Clojure and I would be very surprised if this > wouldn't work the same way in ClojureScript. If such radically different > behaviour is to be expected, then I understood something seriously wrong > about ClojureScript. > > In particular it kills sharing of "clojure only" libraries between both > target platforms. > > Sincerely > Meikel > > > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
On Thu, Sep 22, 2011 at 11:12 AM, Michael Fogus wrote: > > the body of loop in a function to preserve bindings if other fns are > > discovered that close over them so as to avoid introducing a perf hit. > > It would be interesting to see if GClosure is smart enough to deal > with the naive implementation. I know what I'll experiment on > tomorrow. :-) On further consideration I think it's more desirable to wrap the fns at their declaration site in a function which is immediately called to preserve the loop bindings. No need for the entire loop body to pay for preserving locals. David -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
> the body of loop in a function to preserve bindings if other fns are > discovered that close over them so as to avoid introducing a perf hit. It would be interesting to see if GClosure is smart enough to deal with the naive implementation. I know what I'll experiment on tomorrow. :-) -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
There's a patch there but I think it's too simplistic. It should only wrap
the body of loop in a function to preserve bindings if other fns are
discovered that close over them so as to avoid introducing a perf hit.
David
On Thu, Sep 22, 2011 at 8:46 AM, David Nolen wrote:
> It's a known issue:
>
> http://dev.clojure.org/jira/browse/CLJS-39
>
> David
>
> On Wed, Sep 21, 2011 at 2:34 PM, Eric Harris-Braun <
> zippy.314@gmail.com> wrote:
>
>> Check out this little bit of code:
>>
>> (doseq [hid ["a" "b" "c"]]
>> (goog.dom.appendChild (goog.dom.$ "some-element-id")
>> (goog.dom.createDom "div" (.strobj {"id" hid}) (str "Test-"hid)))
>> (goog.events.listen (goog.dom.$ hid) goog.events.EventType.CLICK,
>> (fn [e] (js/alert hid
>>
>> What I want it do to is add in a few div's that when clicked on simply
>> alert with the value their id value. That doesn't actually happen,
>> what happens is that they all alert with the "c" ie. the last value in
>> the list. This code, on the other-hand works:
>>
>> (defn hidfn [hid]
>> (fn [e] (js/alert hid))
>> )
>> (doseq [hid ["a" "b" "c"]]
>> (goog.dom.appendChild (goog.dom.$ "the-receptor")
>> (goog.dom.createDom "div" (.strobj {"id" hid}) (str "Test-"hid)))
>> (goog.events.listen (goog.dom.$ hid) goog.events.EventType.CLICK,
>> (hidfn hid)))
>>
>> Shouldn't the clojurescript compiler detect that the usage of "hid" in
>> the function in the first case requires the creation of an anonymous
>> function to close around that value? Why do I have to do it manually?
>>
>> --
>> You received this message because you are subscribed to the Google
>> Groups "Clojure" group.
>> To post to this group, send email to clojure@googlegroups.com
>> Note that posts from new members are moderated - please be patient with
>> your first post.
>> To unsubscribe from this group, send email to
>> clojure+unsubscr...@googlegroups.com
>> For more options, visit this group at
>> http://groups.google.com/group/clojure?hl=en
>
>
>
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
On Thu, Sep 22, 2011 at 8:46 AM, David Nolen wrote: > It's a known issue: > http://dev.clojure.org/jira/browse/CLJS-39 I somehow missed that one and had filed this one as well: http://dev.clojure.org/jira/browse/CLJS-59 --Chouser -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
It's a known issue:
http://dev.clojure.org/jira/browse/CLJS-39
David
On Wed, Sep 21, 2011 at 2:34 PM, Eric Harris-Braun
wrote:
> Check out this little bit of code:
>
> (doseq [hid ["a" "b" "c"]]
> (goog.dom.appendChild (goog.dom.$ "some-element-id")
> (goog.dom.createDom "div" (.strobj {"id" hid}) (str "Test-"hid)))
> (goog.events.listen (goog.dom.$ hid) goog.events.EventType.CLICK,
> (fn [e] (js/alert hid
>
> What I want it do to is add in a few div's that when clicked on simply
> alert with the value their id value. That doesn't actually happen,
> what happens is that they all alert with the "c" ie. the last value in
> the list. This code, on the other-hand works:
>
> (defn hidfn [hid]
> (fn [e] (js/alert hid))
> )
> (doseq [hid ["a" "b" "c"]]
> (goog.dom.appendChild (goog.dom.$ "the-receptor")
> (goog.dom.createDom "div" (.strobj {"id" hid}) (str "Test-"hid)))
> (goog.events.listen (goog.dom.$ hid) goog.events.EventType.CLICK,
> (hidfn hid)))
>
> Shouldn't the clojurescript compiler detect that the usage of "hid" in
> the function in the first case requires the creation of an anonymous
> function to close around that value? Why do I have to do it manually?
>
> --
> You received this message because you are subscribed to the Google
> Groups "Clojure" group.
> To post to this group, send email to clojure@googlegroups.com
> Note that posts from new members are moderated - please be patient with
> your first post.
> To unsubscribe from this group, send email to
> clojure+unsubscr...@googlegroups.com
> For more options, visit this group at
> http://groups.google.com/group/clojure?hl=en
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
Hi, Disclaimer: I haven't touched ClojureScript a single time up to now, I only know the talk of Rich about it. Neither am I JavaScript developer. Too me this smells like a bug. ClojureScript is not JavaScript. It tries to bring (where possible) Clojure semantics to the JavaScript VM. The first version is perfectly valid Clojure and I would be very surprised if this wouldn't work the same way in ClojureScript. If such radically different behaviour is to be expected, then I understood something seriously wrong about ClojureScript. In particular it kills sharing of "clojure only" libraries between both target platforms. Sincerely Meikel -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
Re: clojurescript closure problem?
No, because that might not be what you want your code to do. As a JavaScript programmer you have to understand what function scope means, and how to use it appropriately. If "hid" is a variable that's in scope when you define an anonymous function, your functions will refer to that specific variable. If you define multiple anonymous functions in the same scope, they will likewise all reference the same specific variable (and sometimes that's what you want). If you want each new anonymous function to have its own separate variable named "hid", then the way you tell the compiler what you want is by wrapping another function around the code that generates your anonymous function, and passing "hid" in as the argument to the wrapper function. Having the compiler try and second-guess what it thinks you *meant* would be bad, because someday some JavaScript programmer would try to share a variable between multiple anonymous functions, and would be scratching their head trying to figure out why the "smart" compiler was failing to obey the rules of JavaScript function scope. m On Wed, Sep 21, 2011 at 2:34 PM, Eric Harris-Braun wrote: > > Shouldn't the clojurescript compiler detect that the usage of "hid" in > the function in the first case requires the creation of an anonymous > function to close around that value? Why do I have to do it manually? > > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
clojurescript closure problem?
Check out this little bit of code:
(doseq [hid ["a" "b" "c"]]
(goog.dom.appendChild (goog.dom.$ "some-element-id")
(goog.dom.createDom "div" (.strobj {"id" hid}) (str "Test-"hid)))
(goog.events.listen (goog.dom.$ hid) goog.events.EventType.CLICK,
(fn [e] (js/alert hid
What I want it do to is add in a few div's that when clicked on simply
alert with the value their id value. That doesn't actually happen,
what happens is that they all alert with the "c" ie. the last value in
the list. This code, on the other-hand works:
(defn hidfn [hid]
(fn [e] (js/alert hid))
)
(doseq [hid ["a" "b" "c"]]
(goog.dom.appendChild (goog.dom.$ "the-receptor")
(goog.dom.createDom "div" (.strobj {"id" hid}) (str "Test-"hid)))
(goog.events.listen (goog.dom.$ hid) goog.events.EventType.CLICK,
(hidfn hid)))
Shouldn't the clojurescript compiler detect that the usage of "hid" in
the function in the first case requires the creation of an anonymous
function to close around that value? Why do I have to do it manually?
--
You received this message because you are subscribed to the Google
Groups "Clojure" group.
To post to this group, send email to clojure@googlegroups.com
Note that posts from new members are moderated - please be patient with your
first post.
To unsubscribe from this group, send email to
clojure+unsubscr...@googlegroups.com
For more options, visit this group at
http://groups.google.com/group/clojure?hl=en
