My GSoC project [Was: Re: DevelopInUFRaw extension]

[Ruben Vermeersch <ruben savanne be>]

On Sun, 2009-05-10 at 19:37 +0300, Anton Keks wrote:
> Hey Ruben!
> 
> Maybe you would like to share the content of your accepted proposal
> with the community?
> It's the same old bazaar vs cathedral topic :-)
> 
> The link below is inaccessible for majority of people here,
> unfortunately.
> Moreover, do you (or will you) commit your work to some public branch,
> so we could follow your work?
> 
> Please keep in touch,
> Anton
> 
> On Thu, May 7, 2009 at 5:45 PM, Pierre-Luc Beaudoin
> <pierre-luc pierlux com> wrote:
>
>         There is a Summer of Code planning to add RAW photography to
>         F-Spot:
>         http://socghop.appspot.com/student_proposal/review/google/gsoc2009/rubenv/t123844493107
>         
>         You should get in touch with him.  He's in CC.


Oh yes, I have absolutely no problem with that, I'm more of a bazaar
person (though when it comes to version control, I prefer git).

Here it is:

Abstract: 

        This Summer of Code proposal is all about finally getting the
        value out of that DSLR which you paid so much for. It's about
        finally bringing RAW support to F-Spot. 
        
Content:

        Support for RAW files [1] in F-Spot is currently nearly
        nonexistant. There's the "Develop in UFRaw"-plugin, but all that
        lets you do is start UFRaw from within F-Spot and import the
        generated JPG back into F-Spot.
        
        This is very cumbersome (you have to start UFRaw for every image and wait
        until it has finished processing) and leads to a lot of problems due to
        missing integration. Ever tried to upload a RAW file to Flickr? Don't forget
        to develop it in UFRaw first, or it will fail!
        
        Compare this to the functionality of e.g. Adobe Lightroom [2] or Apple
        Aperture [3] and you'll quickly notice that we are seriously lagging behind.
        Photographers using free software have the right to be productive too, lets
        enable it!
        
        As such, I propose to bring RAW processing to F-Spot, bringing us repeatable
        editing and first class RAW management. 
        
        This will be based on dcraw (as libopenraw is far from ready, nor does it
        appear that it will be anytime soon), yet should be written in such a way that
        it's possible to transition to something better, once it pops up.
        
        For now it seems wisest to keep editing information inside F-Spot. Once The
        Gimp gains repeatable editing (which will not happen in the timeframe of this
        years GSOC), it might be a good idea to standardise on a shared way of
        expressing RAW processing information. This will allow seamless integration,
        where initial processing can be done in F-Spot and further touchups in The
        Gimp. 
        
        Proposed Mentor: Stephane Delcroix
        
        [1]: http://en.wikipedia.org/wiki/Raw_image_format
        [2]: http://www.adobe.com/products/photoshoplightroom/
        [3]: http://www.apple.com/aperture/
        
        Deliverables
        ============
        
         * A new PhotoView, which offers progressive rendering, allowing for faster
           image displaying and the possibility to process large RAW files quickly.
        
         * RAW Processing and Editing integrated into F-Spot itself, allowing for
           repeatable editing. This should at least integrate Exposure Compensation
           and Color Temperature, preferably Curves and all the rest too.
        
         * RAW files as first class citizens in F-Spot, rather than forcefully
           generating JPG files and being done with it.
        

This is all subject to how well it goes. Am still evaluating whether
we'll use dcraw, libraw or some third alternative [1].

Regarding the public branch: yes. That's why I'm so happy we have git
now. It should all be live and testable. If possible I want to work
directly on master, so that others can contribute too, but we'll have to
evaluate how fargoing the changes will be. Certainly in a branch at
least.

So it's all about good RAW support. Do check out the bits about working
together with gimp. I think this is the strategy we should go for with
f-spot and I'm very interested in your thoughts about this.

So everybody: what is your idea of where we should take f-spot? And
which parts of that vision are you willing to help implement?

Cheers,
   Ruben



[1] One that's mature and active. As much as I like the idea of the
project, these two properties don't apply to libopenraw yet.


--
Ruben Vermeersch (rubenv)
http://www.savanne.be/