[IPython-dev] Moving forward with GUI's (was Re: Help! TR: TR: Ipython plugin

[Barry Wark]

On Tue, Jun 3, 2008 at 1:39 PM, Ville M. Vainio <vivainio at gmail.com> wrote:
> On Tue, Jun 3, 2008 at 6:10 AM, Barry Wark <barrywark at gmail.com> wrote:
>
>> I would encourage you guys to take a look at the
>> ipython1.frontends.frontendbase class. My intention is to develop this
>> into a base class for all GUI frontends. The underlying assumption is
>
> It doesn't seem to be too hard to shoehorn the frontendbase between
> the gui and ipython; most of the work is really in implementation of
> the actual gui drawing.

Correct.  This is the only way to make sure that a terminal frontend,
a graphical frontend like Mathematica's or a web-based frontend can
all fit into the same framework. This leads to the somewhat simplified
M(VC) picture where the FrontendBase acts as the model for an ipython
engine, and the front end implementation acts as a combined view and
controller in the MVC architecture.

>
> Though, at the moment frontendbase can't be directly imported in core
> ipython due to twisted dependency. I suppose the only real twisted
> dependency in that class could be removed by creating
> "result_ready(r)" method and implementing execute_with_callback in
> engine that does normal callback instead of returning a deferred...
> obviously the engine can use deferred for actual implementation, for
> the twisted version.

Because FrontendBase.__init__ takes an engine parameter, the terminal
frontend could use core instead of engineservice for its engine. The
only requirement then is to have a small amount of code (I think this
already exists somewhere in ipython1.core) for chaining callbacks to
the execution result (i.e. creating a pseudo-Deferred).


>
> --
> Ville M. Vainio - vivainio.googlepages.com
> blog=360.yahoo.com/villevainio - g[mail | talk]='vivainio'
>