Re: [gwt-contrib] Re: GWTCon 2015 keynote question

[Stephen Haberman]

>
> The important thing is to keep an open mind, not trying to force JS
> frameworks into our current best practices, and willingness to be inspired
> by the best practices of the chosen JS framework.
>
> Our best practices has emerged from the constraints of GWT, and those
> constraints has just changed.
>

Good points. :-)

- Stephen

-- 
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 google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CANnRKuUVrcwg%3Dx%3DFo5BE1Pk79Dk8bmQmDbvj-KMPXegJUmwT5A%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

Re: [gwt-contrib] Re: GWTCon 2015 keynote question

[Brian Pedersen]

>
> Yeah. The conundrum for me is that I'm addicted to the pure-JVM unit 
> testing/debugging that MVP gives you.
>

I believe Angular controllers, Flux Stores etc. can be just as 
JUnit-testable as MVP, when implemented in Java and exposed through 
JsInterop.

There is a lot of pioneering to do, and I can imagine future best practices 
varying based on the JS framework of choice, as well as other parameters.

The important thing is to keep an open mind, not trying to force JS 
frameworks into our current best practices, and willingness to be inspired 
by the best practices of the chosen JS framework.

Our best practices has emerged from the constraints of GWT, and those 
constraints has just changed.

At least that is my point of view :)

-- 
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 google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/dd951ccd-3764-4bc8-b3c8-16c9a47fd619%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Re: [gwt-contrib] Re: GWTCon 2015 keynote question

[Stephen Haberman]

This is an interesting question (what to use for a greenfield GWT
project)...


> UiBinder: Would need a rewrite to use APT. I guess no one has looked into
> it yet. One main issue is that APT won't get triggered when updating
> resources (*.ui.xml)
>

Yeah, I would probably skip APT and just use an old-school "invoke it
before you run javac" code generator for ui.xml files (and hooked into
run-on-save with an Eclipse builder).

Given how egregiously statically typed UiBinder files are (in a good way),
ideally you wouldn't need much/any reflection/type oracle-type information
(which is where being in an APT environment is nice).

I might eventually prototype something like this, for migrating our
Tessell-based app off widgets...

So the absolute safest solution is to use JsInterop + any JavaScript
> framework. If that framework uses HTML templates then you still have that
> UiBinder feeling. You are not required to use Angular, but its a nice fit.
> The next best solution is probably GQuery and/or Elemento but you should
> expect some refactor work.
>

Yeah. The conundrum for me is that I'm addicted to the pure-JVM unit
testing/debugging that MVP gives you.

So, I guess you could still do MVP, with a view interface, and then have
the view implementation use Angular/JSInterop? That would be
interesting...not sure how it'd work in practice though, as it seems like
none of the watch/scope/etc. logic that you see in Angular controllers
would work in a pure-JVM unit test?

My worry about "just pick a mainstream JS framework and use it via
JSInterop" is that if you're a) coupled to a JS environment for unit
testing and b) interfacing with a framework that is inherently
dynamic/untyped, what's the benefit of using GWT in the first place?

Perhaps this is a naive question, and some GWT+JSInterop+JS framework
examples would (already?) show that you can leverage static typing enough
to make GWT worthwhile in a traditionally non-GWT/JS framework.

It will be interesting to see what Singular looks like. Personally, I think
there is definitely a space to be filled for a next-gen/J2CL-compatible GWT
framework, but I think it'll be ~6-12 months before we really know if a
solid choice materializes. And hopefully we'll have a few choices by then
as well.

Granted, that is not great advice for what to do right now...

- Stephen




>
>
> -- J.
>
> --
> 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 google-web-toolkit-contributors+unsubscr...@googlegroups.com.
> To view this discussion on the web visit
> https://groups.google.com/d/msgid/google-web-toolkit-contributors/d92e7eec-fe04-48fa-ba9a-bcf4c5ed5620%40googlegroups.com
> 
> .
>
> For more options, visit https://groups.google.com/d/optout.
>

-- 
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 google-web-toolkit-contributors+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-web-toolkit-contributors/CANnRKuX3D84j7xP7MXFo7gd7CHh3DtpTADGbhucmyUWMyj8UBA%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.