Re: GSOC 2011

[Simon Lindgren <simon n lindgren gmail com>]

ons 2011-03-23 klockan 12:06 -0500 skrev W. Michael Petullo:
> As you say, DPAP is typically used on a LAN, primarily because of it's
> integration with mDNS. From you additional description, it seems you would
> need either a solution that is more F-Spot specific (than DPAP) or a
> general-purpose home directory synchronization technique.
> 

Okay, so the basic ways to go are:
1. Store all metadata inside the image files or as sidecars, and then
synchronize the picture folder on a file-by-file basis.
2. Create a specialized server software to handle the metadata syncing
and file transfers.

Option one could be tricky to use for implementing some features, like
adding tags and new versions of photos but it does not require a
separate server installation, merely a network filesystem.

I think option two would be more desirable from an extensibility
standpoint, and it could also have some other potential benefits.

So, currently, this is what I think the deliverables of this project
would be:
 * Well written protocol specification.
 * Server implementation of this protocol.
 * F-Spot plugin to use the protocol and the server.
 * (If needed) F-Spot refactorings to support the plugin.

The requirements of the protocol would be the ones stated in my previous
mail (re-posted and slightly rephrased here for clarity):
 * Support the full range of f-spot metadata (including data supplied by
plugins (if that is supported by F-Spot)).
 * Support the full range of operations available inside f-spot (ie.
create new version, add images, add tags, ...)
 * Support downloading pre-generated thumbnails for a set of images.

That last requirement may not be very useful, because F-Spot might need
refactorings before it can download the images on demand, ie the plugin
might have to ensure that all images exist on the drive all the time.

So, are there any thoughts on this? Is it a worthwhile project?

-- 
Simon Lindgren