These are chat archives for canjs/canjs

11th
Apr 2016
Justin Meyer
@justinbmeyer
Apr 11 2016 00:47
can-fixture has been updated: https://github.com/canjs/can-fixture
Quick survey ... anyone like docs in the github repo? Instead of something like: https://canjs.com/docs/can.fixture.html
Thomas Sieverding
@Bajix
Apr 11 2016 00:47
@justinbmeyer Does that have a fix for custom ID mapping?
Justin Meyer
@justinbmeyer
Apr 11 2016 00:47
for the store?
Thomas Sieverding
@Bajix
Apr 11 2016 00:47
Yea
Justin Meyer
@justinbmeyer
Apr 11 2016 00:48
yes
you have to use the set algebra
but it works via that
Thomas Sieverding
@Bajix
Apr 11 2016 00:48
Previously can.Model.id has never played nice with fixtures
Justin Meyer
@justinbmeyer
Apr 11 2016 00:48
canjs/can-fixture#12
Thomas Sieverding
@Bajix
Apr 11 2016 00:48
nice
Justin Meyer
@justinbmeyer
Apr 11 2016 00:49
// Describe the services parameters:
var todoAlgebra = new set.Algebra({
    set.comparators.id("_id"),
    set.comparators.boolean("completed"),
    set.comparators.rangeInclusive("start","end"),
    set.comparators.sort("orderBy"),
});

// Create a store:
var todoStore = fixture.store([
    {
        _id : 1,
        name : 'Do the dishes',
        complete: true
    }, {
        _id : 2,
        name : 'Walk the dog',
        complete: false
    }],
    todoAlgebra );

// Hookup urls to the store:
fixture("/todos/{_id}", todoStore);
Thomas Sieverding
@Bajix
Apr 11 2016 00:50
Ohhh does this play nicely with MongoBD now?
Would you do something like set.comparators.sort(‘$sort’) play nicely w/ compound sorts?
Justin Meyer
@justinbmeyer
Apr 11 2016 00:54
You'd have to make your own sort comparator probably
Thomas Sieverding
@Bajix
Apr 11 2016 00:54
Ah
Justin Meyer
@justinbmeyer
Apr 11 2016 00:54
the one that exists now by default only sorts on one column
you'd mostly need something like that
sortPropValue would be the value of $sort
$sort=______VALUE____
Thomas Sieverding
@Bajix
Apr 11 2016 00:55
I’ve been meaning to update this and add in some more documentation, but I put together a decent lib a couple of months ago that basically wraps mongoose models, adds some hooks
Query params === MongoDB operators
It would be cool if we could do a sample app that married the two
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 09:09
@justinbmeyer for the can-fixture docs servey, what you mean? docs will be in raw files?
@justinbmeyer I dont like the docs be in github
Dovid Bleier
@dbleier
Apr 11 2016 14:19
@justinbmeyer I think there is value in having the docs in both places in the different formats. Sometimes it is easier to read in one format and sometimes in the other.
I have the following stache
<can-import from="cms/sign-controls/" />
<sign-controls {^selectedPresentation}="*presentation"></sign-controls>

<can-import from="cms/presentation-workspace/" />
<presentation-workspace {presentation}="*presentation"></presentation-workspace>
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 14:21
@dbleier hi, just now I remember I have to check your issue sorry :/
Dovid Bleier
@dbleier
Apr 11 2016 14:21
in sign-control.js this DOES get called and the prop set
    selectPresentation(p) {
        this.attr('selectedPresentation', p);
    },
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 14:21
I wish it was fixed for you
Dovid Bleier
@dbleier
Apr 11 2016 14:21
@cherifGsoul np, thanks for looking into it
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 14:22
not fixed yet?
Dovid Bleier
@dbleier
Apr 11 2016 14:22
but in the presentation-workspace.js the presentation prop is NOT being set. And this line is not called
    presentation: {
        set(newval) {
            console.log('Loaded Presentation', newval);
            debugger;
            return newval;
        }
    }
@cherifGsoul no, not yet. I put that project on hold for a bit. It could be using stache vars there might be a way to get back into the VM after fall-through-cache updates, but I haven't tried it yet
in relation to my current question, any one see what is wrong or have an idea?
Kevin Phillips
@phillipskevin
Apr 11 2016 14:24
should {^selectedPresentation} be kebab case?
{^selected-presentation}
Dovid Bleier
@dbleier
Apr 11 2016 14:25
even though that is how it is in the VM?
Kevin Phillips
@phillipskevin
Apr 11 2016 14:25
right
html attributes always have to be like that I think
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 14:26
for attrs kebab case
Kevin Phillips
@phillipskevin
Apr 11 2016 14:26
html is case insensitive
Dovid Bleier
@dbleier
Apr 11 2016 14:26
aha! that worked
I am going to post this on the forums
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 14:27
yes please
Kevin Phillips
@phillipskevin
Apr 11 2016 14:27
sounds good
Kevin Phillips
@phillipskevin
Apr 11 2016 14:38
:thumbsup:
Mohamed Cherif Bouchelaghem
@cherifGsoul
Apr 11 2016 14:38
thank you
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:07
is there a way to remove a component from the list of available components?
Matthew Phillips
@matthewp
Apr 11 2016 18:24
what do you mean?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:25
I am running unit tests. One unit test run can.component.extend({tag: ‘xxx’, …});
In another unit test I don’t want component with tag ‘xxx’ to be expanded.
I can’t control the order in which tests run
Matthew Phillips
@matthewp
Apr 11 2016 18:26
unofficially you can delete it from can.view.callbacks._tags
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:26
thank-you very much
Matthew Phillips
@matthewp
Apr 11 2016 18:27
i would suggest creating distinct tag names in tests though
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:27
I am testing components that are being used in the final code
I added can.view.callbacks._tags = {};, but my tags are still being expanded
Matthew Phillips
@matthewp
Apr 11 2016 18:31
what do you mean by "being expanded"?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:32
I was expecting <xxx></xxx> instead the html was 3000 characters long.
however, if I run just the one file in my test I get the expected results
Matthew Phillips
@matthewp
Apr 11 2016 18:34
you want to deregister the component so that when it is rendered it is treated as an unknown element, correct?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:34
yes - looking at the code I can see that I will need to deregister each component individually
Matthew Phillips
@matthewp
Apr 11 2016 18:35
it's unusually to do can.Component.extend in a test, is there perhaps a better way to achieve what you are trying to test?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:36
I will consider that
Matthew Phillips
@matthewp
Apr 11 2016 18:36
what is it you are trying to do? I can maybe think of an idea
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:37
I am trying to test as much of the code as possible - code coverage is a big deal at my company
        for (var tag in can.view.callbacks._tags) {
            if (can.view.callbacks._tags.hasOwnProperty(tag)) {
                delete can.view.callbacks._tags[tag];
            }
        }
That successfully clears registered components
Matthew Phillips
@matthewp
Apr 11 2016 18:41
code coverage is definitely good
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:41
I have components and some have event code. I want to write unit tests for each of the events in the components. What I have done is to run the can.Component code, create a stache fragment that uses the component tag and add the stache fragment to the DOM. I then write unit tests to trigger each of the events within the component.
Matthew Phillips
@matthewp
Apr 11 2016 18:42
are these real components that your app uses or are you dynamically creating components for the sake of testing events?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:42
Real components
Matthew Phillips
@matthewp
Apr 11 2016 18:42
hm, then I don't get it
I don't get the deletion part
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:42
The test of the component works just fine
Matthew Phillips
@matthewp
Apr 11 2016 18:42
every time you render a template it creates a new instance of the component
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:43
I then have a test for the code that creates the stache fragment used in production. In writing tests for that code, I don’t want the component expanded.
As an aside, what happened to the PR I created for exclusiveMapping?
Guido Smeets
@gsmeets
Apr 11 2016 18:45
why would you want to test your unregistered component not rendering?
seems a bit over the top
Matthew Phillips
@matthewp
Apr 11 2016 18:46
I think I follow, why don't you want the component expanded though
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:46
I want to get the same results if I run the test by itself or if it runs as part of the whole unit test run.
Matthew Phillips
@matthewp
Apr 11 2016 18:47
Ok, so why wouldn't you get the same results?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:47
If changes are made to the components or stache files then the test for the other code will fail.
In the unit test for the base code, the code that creates the stache fragment, I didn’t include the files that generate the components. When it runs alone I get one set of results. When I run it in combination with other tests, I get different results.
Matthew Phillips
@matthewp
Apr 11 2016 18:49
Ok, so i'm starting to understand
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:49
One of my tests is the length of string created by the stache fragment
Matthew Phillips
@matthewp
Apr 11 2016 18:49
The code that creates the stache fragment doesn't depend on all of the components
This is your app's main, right?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:50
No it doesn’t, in fact I write mocks to bypass all component and viewModel methods
It’s the piece of main code of this section of the system
I am replacing a small piece of a large legacy system with code using CanJS
Matthew Phillips
@matthewp
Apr 11 2016 18:51
So if this code is rendering a fragment expecting some components to be registered, it should depend on those components
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:52
That’s not what I am testing when writing unit tests for the base code. And I don’t want the base code unit tests to break if someone makes a change to one of the components.
Matthew Phillips
@matthewp
Apr 11 2016 18:52
yes you do!
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:53
The base won’t care if someone adds a class to some sub-component for the purpose of changing how styling works
Kevin Phillips
@phillipskevin
Apr 11 2016 18:54
so you really just want to test what can.stache gets called with? and then what jQuery.append gets called with?
            var stache = can.stache(‘<xxx id="' + id + '" {parent}="parent"></xxx>');
            this.$container.append(stache({
                parent: this
            }));
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:55
Yes, but I want the append call to be unexpanded.
Kevin Phillips
@phillipskevin
Apr 11 2016 18:55
you’re testing this kind of thing?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:55
yes
Kevin Phillips
@phillipskevin
Apr 11 2016 18:55
so your test is like expect($container.html()).toBe(‘<xxx id=“5”></xxx>’); ?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:56
yes
Matthew Phillips
@matthewp
Apr 11 2016 18:56
ah
Kevin Phillips
@phillipskevin
Apr 11 2016 18:56
are you using steal?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:57
no. This is a replacement of a tiny piece of a large legacy system that doesn’t use Steal
Kevin Phillips
@phillipskevin
Apr 11 2016 18:57
hmm, ok
one way to do it is to create a module that does the can.stache call and the append
then spy on that function for this test
Christopher Oliphant
@RALifeCoach
Apr 11 2016 18:58
I can clear the registered components using:
        for (var tag in can.view.callbacks._tags) {
            if (can.view.callbacks._tags.hasOwnProperty(tag)) {
                delete can.view.callbacks._tags[tag];
            }
        }
I have tried that and my tests are working whether they run individually or as part of the whole test run
Kevin Phillips
@phillipskevin
Apr 11 2016 19:00
ok, so is something still not working?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:01
Everything is working. Matthew felt there was a better way of writing the tests. We were discussing options.
Kevin Phillips
@phillipskevin
Apr 11 2016 19:01
ok, sorry to make you recap
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:02
At this point in time I don’t see a better way and I will stick with this.
What ever happened to my PR - the one for exclusiveMapping?
No problem
You helped Matthew understand what I was trying to do.
Kevin Phillips
@phillipskevin
Apr 11 2016 19:08

only other way I can think of (not sure if it’s better) is create a module for it

// domHelpers.js
module.exports = {
    append: function($container, componentString, data) {
            var stache = can.stache(componentString);
            this.$container.append(stache(data));
    }
}

then in your tests you can do

var stub = sinon.stub(domHelpers, ‘append’);

expect(stub.getCall(0).args[1]).toEqual(‘<xxx id=“5”></xxx>’);
might be better if you’re calling append from a lot of different places
shorter/faster than clearing out all the tags every time
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:10
okay. At this point in time there is only the one append, but I could still do that.
Kevin Phillips
@phillipskevin
Apr 11 2016 19:11
it’s another option. up to you how you want to do it.
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:11
Thanks
Kevin Phillips
@phillipskevin
Apr 11 2016 19:12
if you were using steal to load your tests you could use steal-clone to make this easier
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:12
I don’t have the time to change the existing 6500+ unit tests written so far
Kevin Phillips
@phillipskevin
Apr 11 2016 19:12
I hear you
that’s a ton of tests
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:14
We have a massive project underway to add unit testing to this legacy code base. We are approaching 80%. I suspect we will be over 8000 by the time this project is complete.
Kevin Phillips
@phillipskevin
Apr 11 2016 19:14
is it possible to run subsets of the tests? or do you have to run them all every time?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:15
jasmine lets me select specific tests to run by changing describe to fdescribe, but no, the tests were not split into sub-sections
that would have made sense, given the size, but that is a lesson learned
Kevin Phillips
@phillipskevin
Apr 11 2016 19:17
ok
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:17
Any news on the PR's I created?
Kevin Phillips
@phillipskevin
Apr 11 2016 19:19
I’m not sure
probably want to talk to @justinbmeyer
Matthew Phillips
@matthewp
Apr 11 2016 19:19
you should reach out to @daffl or @justinbmeyer
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:19
okay - will do, thanks
Matthew Phillips
@matthewp
Apr 11 2016 19:20
it's ok to ping them in the issues every once in a while, everyone is busy :)
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:20
Alright - I will do that. You guys have produced an amazing toolset - especially while working full time jobs.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:21
@RALifeCoach I think your best bet is to figure out a way to weave this into CanJS
we can't add it until 2.4 (or 3.0) because it's a new feature
(btw, it seems to be breaking the build)
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:21
@justinbmeyer I wove the change into map/define, if you are referring to the PR
Justin Meyer
@justinbmeyer
Apr 11 2016 19:22
I am ... I'm saying weave it in w/o needing to change CanJS
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:23
I can fix the build breaking - I suspect there are some test cases that broke
Right now, I have modified the can.custom file for our site.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:24
Can you remind me why this wasn't done w/ parseModel?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:24
I did have this code outside of CanJS, it just seemed like a feature other people may want to use.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:25
it likely is something a certain percentage of people might want ... similar to something like can/map/backup
and the define plugin itself
but we typically try to not change code to add these things
or, at minimum, add some hooks
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:26
I like the idea of maps. They tell everyone what to expect. They provide clear and consice documentation built right into the code. But then I could pass anything else into the map and it would just accept it without complaint.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:26
that allow the other code to "weave" itself in
hopefully I'm making sense
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:27
I originally created a new file can/xdefine or something. But I had to duplicate most of the can/define code.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:27
yeah, there's probably a way to avoid duplicating the code
newValue.parent = this; probably might cause a memory leak
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:28
I can have another look about the duplicated code, but can/define uses several private methods.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:29
actually probably not a memory leak, but a bug
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:29
Fair comment on parent = this. I can remove it.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:29
if anyone uses parent
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:29
I can specifically pass it when I need it
Justin Meyer
@justinbmeyer
Apr 11 2016 19:29
can/define uses private methods ... this is why I said "or, at minimum, add some hooks"
we'll expose some things so other plugins can do what they need
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:30
okay - I can look at minimum changes to define to provide the hooks
Justin Meyer
@justinbmeyer
Apr 11 2016 19:30
but that's different than having can/map provide every bit of functionality straight away
what we can do is create a can-map-exclusive project too.
are you using StealJS to load modules from npm?
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:31
The other side effect from my changes, that I found very useful, is that it lets me order properties.
That can be useful when writing setters that rely on other properties being set.
No I am not using StealJS to load modules from npm
I am doing all of my testing by modifying my version of can.custom and then replicating the changes into the can js code
Justin Meyer
@justinbmeyer
Apr 11 2016 19:34
hmmm
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:34
Not the most foolproof way of making the changes.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:34
yeah, we only merge things in once they have tests
in an ideal situation, you'd be familiar with https://donejs.com/plugin.html
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:34
At this point in time, what I wanted from you folks was either - go ahead we like the idea or no we will never merge it because we hate the idea.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:35
ok, so cutting to the chase then .... we will never merge this, not because I hate the idea though
well, I would merge some hooks
to make it possible
however, adding the feature (which I don't hate) would mean adding a feature
which would mean going to 2.4
or making it part of 3.0 which is our next release
now for 3.0, we will have can-define which effectively should solve this ... especially in combination with parseModel
so this improvement is sorta stuck in limbo
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:37
okay - so if I made the changes to add hooks in a way that doesn’t change the existing functionality and then wrote another plugin, then you would consider merging the code and listing the plugin
Justin Meyer
@justinbmeyer
Apr 11 2016 19:37
absolutely
like I was saying, in an ideal situation, you could quickly make the plugin with: https://donejs.com/plugin.html
then we'd tweet about it, etc .... in fact, for 3.0, we're basically making everything a plugin of sorts
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:38
Okay - I will spend some time on it and resubmit the PR.
I will also review that link and read up on plug ins
Justin Meyer
@justinbmeyer
Apr 11 2016 19:38
we're going to have a list of our A+ ones (can/stache, can/list, can/compute, can/component), our B ones (can/map/backup), and hopefully a bunch of community ones
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:39
what will define look like in 3.0? will it provide the functionality I am looking for?
Justin Meyer
@justinbmeyer
Apr 11 2016 19:39
our idea is to nurture the community ones so they can eventually be part of CanJS's A+ stuff
can-define is going to be the replacement to can.Map and can/map/define
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:40
Will I have the option to say, “Only include the properties I have listed in the define and ignore all other properties”?
Justin Meyer
@justinbmeyer
Apr 11 2016 19:40
we can likely add it ... it already only makes observable the properties you have listed in the define
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:40
Nice
Justin Meyer
@justinbmeyer
Apr 11 2016 19:40
I've been thinking about if we can make it error if any other properties are added
with Object.defineProperty
but it's 3-4 X faster than can.Map
and probably 5 X faster than can.Map with can/map/define
shows what using it looks like
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:42
My problem was I had an object that had hundreds of properties, but the stache template only needed 30 or so. Being able to say ignore undefined properties meant that CanJS didn’t have to lug the rest of the properties around.
Nice, can.Map is slow.
Justin Meyer
@justinbmeyer
Apr 11 2016 19:42
yeah, typically people would remove those in Model.parseModel
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:43
Wouldn’t they have to write code to remove them? Why ask them to write new code when they have already written maps?
Justin Meyer
@justinbmeyer
Apr 11 2016 19:44
in general, w/ that many properties, they didn't define each one anyway
so new code was necessary no matter what
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:44
that seems like duplication. And if a change now needs one of the properties, I would have to change parseModel and the map
with my change, anything not in a map is ignored by can.Map
Justin Meyer
@justinbmeyer
Apr 11 2016 19:45
and everything is fastest there, while you are dealing with the raw response data
yeah, your change is nicer
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:46
perhaps you could have another look at the PR and see about copying some of the code into version 3.0
Justin Meyer
@justinbmeyer
Apr 11 2016 19:49
can-define works totally different
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:49
okay
Justin Meyer
@justinbmeyer
Apr 11 2016 19:50
I know .. it sucks .. you're kinda in a transition spot
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:50
when is 3.0 due?
Justin Meyer
@justinbmeyer
Apr 11 2016 19:50
probably will have betas out in 2 months
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:50
we could live with the patched can.custom file for now
Justin Meyer
@justinbmeyer
Apr 11 2016 19:51
I'm hoping to work on can-define later this week
finish making it work with can.Component and all the view bindings
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:51
unless you expect a version 2.4 that would make sense to have in my project
Justin Meyer
@justinbmeyer
Apr 11 2016 19:51
I doubt it. 3.0 is already underway
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:51
okay, then that’s probably what I will do
Justin Meyer
@justinbmeyer
Apr 11 2016 19:52
for what you need
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:52
then I can test the performance against the 3.0 code and I may not need the changes
Justin Meyer
@justinbmeyer
Apr 11 2016 19:52
well, I would still make a pull request to open up can.Map
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:52
okay
Justin Meyer
@justinbmeyer
Apr 11 2016 19:52
well, you would have to rewrite some code for can-define
it removes the need for .attr()
so, when 3.0 lands
you'll still be using it's can.Map likely
so it's probably worth it to land a pull request that exposes things
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:53
okay - I will do that
Justin Meyer
@justinbmeyer
Apr 11 2016 19:53
that way your plugin could work against 2.3.23 and 3.0.0
Christopher Oliphant
@RALifeCoach
Apr 11 2016 19:53
alright
thanks for the chat, I’ll let you get back to work. I know you guys are crazy busy.