+1 for the observer pattern. Having the ability to install "middlewares"
within the flow would be great. In this case I would suggest to consider
the opportunity to choose the plugin execution order (i.e. plugin1 request
processing should happen before plugin2, etc.).

+0 in case of the 404 method is employed. In this case I prefer to use an
app out of qgis server...
Il 04/ott/2014 14:52 "Alessandro Pasotti" <[email protected]> ha scritto:

> 2014-10-04 13:17 GMT+02:00 Andrea Peri <[email protected]>:
> > The slides dont say nothing about the optionality of this.
> >
> > Should be possible to disable all the python side using some compile
> settings ?
>
> Of course yes.
>
> > like
> >
> > "disable-python-server-side" ?
> >
> > Or enable it only for speciic project.
>
> This could be easily implemented at the webserver level or directly
> inside the plugins in case of need.
>
> >
> > So avoid to have more insances.
> > Something with python capability and other without python capability ?
> >
> > Aso a last question is the accounting.
>
> The plugin support has no direct implications with authentication or
> authorization. What the plugins will implement (and how they'll do it)
> is far beyond the scope of my RFC.
>
>
> --
> Alessandro Pasotti
> w3:   www.itopen.it
> _______________________________________________
> Qgis-developer mailing list
> [email protected]
> http://lists.osgeo.org/mailman/listinfo/qgis-developer
>
_______________________________________________
Qgis-developer mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/qgis-developer

Reply via email to