incubator-callback-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Julien Bouquillon [revolunet]" <jul...@revolunet.com>
Subject Re: remote platform
Date Mon, 21 May 2012 22:26:21 GMT
Hi Dave,

Nice to know its not such a useless approach ;)

I dont get how this could fit with ripple : some real data from device,
some fakes ?
Or maybe the device could "feed" the ripple interface in realtime ?

Another thing to think about, maybe the "remote" cordova platform could
also be used (in the mid term) to control remote phones, not only for my
debug purposes. (think remote control, video games maybe...)

dnode looks like an awesome lib to transfer JS calls transparently from one
device or browser to another. I have some tests to do with it.

So if we have a "remote" cordova platform, it could be plugged into ripple,
but also maybe as a standalone 'remote' device.

Id love to have some feedback/vision from the cordova/phonegap/ripple team
on these ideas :)

Thanks

Julien (@revolunet)

PS: about the picture, i got the base64 back in the callback and worked
nice. I plan to add a "file" proxy so we can emulate file access...


2012/5/22 Dave Johnson <dave.c.johnson@gmail.com>

> Hi Julien, I agree this sort of thing is needed - I also made a POC
> doing the exact same thing with sockets so that I can write and debug
> a phonegap app in chrome using desktop developer tools but getting the
> actual device data so that I don't need to stub any data nor interact
> with a simulator UI to control various stubbed data. I got stuck with
> some issues around taking pictures and how the paths on the device map
> to paths in the desktop browser, but it's not insurmountable.
>
> I think it makes the most sense to have it work with ripple. For some
> APIs like accelerometer it makes sense to get the data from the device
> (it's easier to shake the device than simulate shake accelerometer
> data) but things like geolocation make sense to be able to use the
> actual geo data from the phone as well as stub data like a route or
> something that you can't do easily with the actual device.
>
> -dave
>
> On Sat, May 19, 2012 at 3:25 AM, Bryan Bishop <kanzure@gmail.com> wrote:
> > On Fri, May 18, 2012 at 11:49 AM, Julien Bouquillon
> > <julien@revolunet.com> wrote:
> >> I have to look at dnode as Bryan pointed out. I may learn a lot from
> this
> >> project and it looks to fit quite well : remote exec, calls
> serialization...
> >
> > I encountered dnode when I was doing some work with phantomjs
> > (headless webkit for scraping and ui testing). Unfortunately,
> > PhantomJS doesn't have many ways to get data in/out of its process,
> > because there are no compiled bindings yet.
> >
> > So, as a consequence, one way to get information out is through
> > javascript on an open WebPage object. In particular, dnode is how the
> > phantomjs-node people use this. See here:
> >
> > https://github.com/sgentle/phantomjs-node#no-really-how-does-it-work
> >
> > Both javascript debugging environments are equally painful :-). I
> > haven't been able to get the phantomjs remote debugging bridge to work
> > yet (I'm pretty sure it does, though), and weinre doesn't.
> >
> > - Bryan
> > http://heybryan.org/
> > 1 512 203 0507
>

Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message