Re: new module proposal: brasero

[Luca Ferretti <elle uca libero it>]

Il giorno lun, 03/11/2008 alle 00.52 +0000, Bastien Nocera ha scritto:

> Problem is it's a stand-alone application. It doesn't integrate with the
> workflow or usage pattern that we set out to achieve with
> nautilus-cd-burner.

Bastien, sorry, but I think you are wrong here. By now Brasero can't be
integrated (or, better, fully integrated) 'cause nautilus-cd-burner is
"hardcoded" here and there in the desktop.

> * Media is inserted first
> - User inserts blank media
> - The CD creator window opens up inviting the user to add files to the
> location to burn onto a CD
> - User clicks "write to CD"
> - nautilus-cd-burner opens, click write, done

Note that in GNOME 2.22 an later you have the ability to choose the
application to launch on media insertion, as well as the ability to
prompt a choice, thanks to Nautilus.

Brasero is well integrated in this workflow (see Nautilus Prefs -> Media
tab -> Other Media group -> Type combobox -> blank XXX disk)

The real issue IMHO is you can't launch brasero from burn:// location or
eternal apps like Rhythmbox.

> Brasero currently only offers an answer to the second option. And what
> we're really looking for is an answer for option 3. I'll take patches to
> create Video DVDs from a Totem playlist, and I'm sure Pitivi hackers
> would be happy for it to have the same treatment.

Let me explain my point of view. The Brasero vs n-c-b fight is showing a
GNOME shortcoming: no flexible way to integrate burning application in
desktop and in applications.

IMHO if we really want a "burning workflow" in GNOME, we need to define
two components:
     1. a "frontend" to add content and adjust the layout 
     2. a "backend" to create the ISO and/or physically move the data on
        discs

Examples of frontends are Rhythmbox playlists, burn:// location in
Nautilus, Brasero main window, i.e. anything will allow you to collect
and organize stuff you want to copy on disks.

Examples of backends are nautilus-cd-burner and brasero :-) OK, serious.
>From UI point of view, backend is the the dialog showing you the
progress of burning process, a software between the frontend and
cd*tools.

Based on this separation between backend and frontend, the optimal
workflow is something like the following: 
     A. open the (frontend) application to collect data to burn
        (manually launched by user or automatically on media insertion)
     B. collect the data using the features provided by current
        application
     C. click "Burn Disk" button
     D. the application sends collected data to burning backend
     E. the burning backend opens a progress dialog and starts to burn

Yes, I know this is exactly how burn:// + n-c-b works, but if we can't
simply switch to burn:// + brasero, I suppose we have an issue in our
framework.

Now, the question is: how many changes do we need in Brasero and other
applications/libraries to achieve the goal of a flexible burning
framework for GNOME?

Sparse Notes:
      * IMHO currently brasero (as "backend") is better then n-c-b, at
        least 'cause brasero provides multisession and
        blanking/formatting
      * About the "frontend" I think could be good put burn:// location
        and brasero side by side in the desktop, having a simple
        Nautilus add-in to quickly burn stuff as well as a full featured
        (but not bloated) application
      * really, can we define a default GNOME burning backend without
        hardconding a command or depending on some library? Any DBUS
        magic?
      * note I'm not suggesting to split brasero in brasero-burner
        (backend) and brasero-cd-creator (fronted) modules/tarballs
      * s/nautilus-cd-burner/gnome-disc-burner