Awesome.

Shaz (or anyone else), curious if you've tested yet to see if the whitelist
still works with WKWebView? (e.g. does it go through NSURLProtocol?)


On Sat, Jun 14, 2014 at 8:16 PM, tommy-carlos williams <to...@devgeeks.org>
wrote:

> This looks promising.
>
> Thanks for the update, Shazron.
>
> - tommy
>
>
> On Sun, Jun 15, 2014 at 7:48 AM, Shazron <shaz...@gmail.com> wrote:
>
>> Rev log: http://trac.webkit.org/log/trunk/Source/WebKit2/
>> UIProcess/API/Cocoa/WKWebView.mm
>> for potential WKWebView updates in beta 2.
>>
>> On Sat, Jun 14, 2014 at 2:46 PM, Shazron <shaz...@gmail.com> wrote:
>>
>>>  Some potential good news. Updated 4 days ago, hopefully its in beta 2:
>>>  http://trac.webkit.org/changeset/169765
>>>
>>>  "Add -[WKWebView evaluateJavaScript:completionHandler:]"
>>>
>>>  Updated: https://issues.apache.org/jira/browse/CB-6884
>>>
>>>
>>>  On Sat, Jun 7, 2014 at 10:06 PM, Carlos Santana <csantan...@gmail.com>
>>> wrote:
>>>
>>>>  ok, was a bit confuse with api doc, I assumed that there was a way to
>>>>  specify a time other than documentstart, documentend, and no passing
>>>>  something will do it immediately.
>>>>
>>>>  Will open a radar too, we need wkwebview to officialy support for
>>>> objc->js,
>>>>  postMessage seems kind of half working if only can do js->objc
>>>>
>>>>
>>>>  On Sat, Jun 7, 2014 at 10:48 PM, Shazron <shaz...@gmail.com> wrote:
>>>>
>>>>   No it's not. That is precisely what we discussed, it's the limitation
>>>>> in
>>>>>  WKUserScript.
>>>>>
>>>>>  On Saturday, June 7, 2014, Carlos Santana <csantan...@gmail.com>
>>>>> wrote:
>>>>>
>>>>>  > Shaz
>>>>>  >   I think the closest replacement is [1] - (void)addUserScript:(
>>>>>  > WKUserScript *)*userScript *
>>>>>  >
>>>>>  > I have not tried my self, but looking forward on helping out.
>>>>>  >
>>>>>  > [1]:
>>>>>  >
>>>>>  >
>>>>>  https://developer.apple.com/library/prerelease/ios/
>>>>> documentation/WebKit/Reference/WKUserContentController_Ref/
>>>>> index.html#//apple_ref/occ/instm/WKUserContentController/addUserScript
>>>>>  > :
>>>>>  >
>>>>>  >
>>>>>  > On Fri, Jun 6, 2014 at 12:47 AM, Shazron <shaz...@gmail.com> wrote:
>>>>>  >
>>>>>  > > No use in polling if we can't write anything back to JS from
>>>>> Obj-C.
>>>>>  > >
>>>>>  > > There's a private API to do so:
>>>>>  > >
>>>>>  > >
>>>>>  >
>>>>>  https://github.com/WebKit/webkit/commit/
>>>>> adb4c60064b38b5ab3d6e78422325f35f0b7fe2b
>>>>>  > > only landed a few months ago, we'll have to do some advocacy
>>>>> through
>>>>>  > > whatever channels we have to get it in the public API (radars,
>>>>>  > > connections), since it is a deficiency in their API losing
>>>>> something
>>>>>  > > like stringByEvaluatingJavaScriptFromString
>>>>>  > >
>>>>>  > >
>>>>>  > >
>>>>>  > > On Thu, Jun 5, 2014 at 6:09 PM, Michal Mocny <mmo...@chromium.org
>>>>> >
>>>>>  > wrote:
>>>>>  > > > Oh wow.  I totally assumed that you can postMessage in either
>>>>>  direction
>>>>>  > > at
>>>>>  > > > any time.  Wouldn't the alternative be polling from JS?
>>>>>  > > >
>>>>>  > > > -Michal
>>>>>  > > >
>>>>>  > > >
>>>>>  > > > On Thu, Jun 5, 2014 at 6:46 PM, Shazron <shaz...@gmail.com>
>>>>> wrote:
>>>>>  > > >
>>>>>  > > >> Well seems like the answer in iOS 8 beta 1 is -- no arbitrary
>>>>>  sending
>>>>>  > > >> of JS, so no Obj-C -> JS communication, which leaves Cordova
>>>>>  > > >> handcuffed. Please everyone file radars for this.
>>>>>  > > >> https://devforums.apple.com/message/975230#975230
>>>>>  > > >>
>>>>>  > > >> On Thu, Jun 5, 2014 at 3:40 PM, Shazron <shaz...@gmail.com>
>>>>> wrote:
>>>>>  > > >> > Thanks Tommy - I sure will.
>>>>>  > > >> >
>>>>>  > > >> > I think injecting JavaScript at arbitrary times -- you would
>>>>> just
>>>>>  > use
>>>>>  > > >> > WKUserScriptInjectionTimeAtDocumentEnd for WKUserScript --
>>>>>  although
>>>>>  > I
>>>>>  > > >> > haven't tested it. If setting JS at arbitrary times is taken
>>>>> away
>>>>>  -
>>>>>  > > >> > yikes.
>>>>>  > > >> >
>>>>>  > > >> > Anyways, on the bridge front, I've posted my approach for
>>>>> the new
>>>>>  > > bridge:
>>>>>  > > >> > https://issues.apache.org/jira/browse/CB-6884
>>>>>  > > >> >
>>>>>  > > >> > On Thu, Jun 5, 2014 at 2:06 PM, Tommy Williams <
>>>>>  to...@devgeeks.org>
>>>>>  > > >> wrote:
>>>>>  > > >> >> I am sure you won't need it, but if I can help, let me know.
>>>>>  > > >> >>
>>>>>  > > >> >> I think the biggest hurdle will be firing user scripts at
>>>>>  arbitrary
>>>>>  > > >> times
>>>>>  > > >> >> instead of only on page load.. There seems to be an API that
>>>>>  hasn't
>>>>>  > > been
>>>>>  > > >> >> exposed :/
>>>>>  > > >> >> On 6 Jun 2014 04:59, "Shazron" <shaz...@gmail.com> wrote:
>>>>>  > > >> >>
>>>>>  > > >> >>> My intent is to work on this today, in a branch for
>>>>> cordova-ios:
>>>>>  > > >> >>> https://issues.apache.org/jira/browse/CB-6863
>>>>>  > > >> >>>
>>>>>  > > >> >>> On Wed, Jun 4, 2014 at 10:15 AM, Shazron <
>>>>> shaz...@gmail.com>
>>>>>  > wrote:
>>>>>  > > >> >>> > Use Safari to watch "Introducing the Modern WebKit API"
>>>>> (no
>>>>>  > login
>>>>>  > > >> >>> required):
>>>>>  > > >> >>> > https://developer.apple.com/videos/wwdc/2014/
>>>>>  > > >> >>> >
>>>>>  > > >> >>> > On Wed, Jun 4, 2014 at 8:25 AM, Michal Mocny <
>>>>>  > mmo...@chromium.org
>>>>>  > > >
>>>>>  > > >> >>> wrote:
>>>>>  > > >> >>> >> You can probably bet on it.
>>>>>  > > >> >>> >>
>>>>>  > > >> >>> >> But this is really fresh news, we're as excited as you
>>>>> are,
>>>>>  > > trying
>>>>>  > > >> to
>>>>>  > > >> >>> >> figure out the details.
>>>>>  > > >> >>> >>
>>>>>  > > >> >>> >>
>>>>>  > > >> >>> >> On Wed, J--
>>>>>  > Carlos Santana
>>>>>  > <csantan...@gmail.com <javascript:;>>
>>>>>  >
>>>>>
>>>>>
>>>>
>>>>
>>>>  --
>>>>  Carlos Santana
>>>>  <csantan...@gmail.com>
>>>>
>>>

Reply via email to