On Wednesday, May 24, 2017 at 4:06:25 AM UTC-4, Thomas Broyer wrote:
>
> Open-sourcing j2cl takes time (see 
> https://opensource.google.com/docs/releasing/ , we've been told it's 
> currently in the "getting approval" phase), more time than for 
> jsinterop-generator or elemental2.
> But no work on GWT 3 will happen before j2cl is open sourced (maybe some 
> of us steering-committee members and core contributors will have access to 
> a private repo so we can start thinking about how to build GWT 3 using it; 
> but no open-source work will likely happen before j2cl is itself open 
> source, and this is specifically why we're still in that "we don't know 
> what GWT 3 will look like" situation).
>

Yes, I understand and am afraid precisely of *all* that.
 

> BTW, for now, the only "hook" that has been identified is a way to get 
> access to private fields (others, like Errai CDI, might want access to 
> private constructors and private methods too).
>
At least something but the form of the hook will influence how well things 
can be done otherwise, see my previous post. 
 

> This could possibly come in the form of a "contract" in j2cl in the 
> "shape" of the ES6 classes it produces and how it represents private 
> members there;
>
and then possibly using jsinterop-base, 
>
That would be (4) from my short list of possibilities. While possible, this 
is less desirable approach than (2) as per my explanation.
 

> or creating a dedicated API based on that contract)
>
Yes, something like that is what I would like. I am not asking J2CL to 
perform any analysis including reachability - I must do it. But I need to 
be able to do it as/when needed. 
 

-- 
You received this message because you are subscribed to the Google Groups "GWT 
Contributors" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/8202e78c-b9f9-488f-bb40-aeaa2b422a71%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Reply via email to