[Deja-dup-hackers] Integrating backup restore into Nautilus



Hey guys,

I tried talking to you on IRC but Cosimo didn't reply and Alex is going on vacation so I thought mailing you would be better for explaining the situation and keeping the rest of my team in loop.

I am a GSOC student working on Deja Dup [0], which is a backup program designed with simplicity in mind so that users without CS/IT degrees (e.g. grandma ;)) can do backups and restore their files in the simplest manner possible.

State of the project basically goes as this: program supports backup and full restore, but during my GSOC time I am trying to address restoration of individual files (partial backup). I already designed a quick restoration of files that were deleted from a certain directory by adding a new menu option to contextual menu. The next issue that I will try to resolve is a bit harder - restoring individual files and/ or directories - basically selecting a subset of files that we have in backup and restoring them.

To do this we thought of several options but semi-current version of mockup that we designed and are trying to implement is located here [1].

One thing that will surely struck you if you look at the mockup is that we will have to use a lot of components that you have implemented for Nautilus and indeed we are re-inventing a file browser with the difference being that we are showing files from backup that are not actually on the file system.

Therefore we naturally thought of developing an extensions for Nautilus but as far as I came with my research this is not something that is possible because Nautilus does not offer any way to add files that don't actually exist on filesystem which is the point at which you guys come in :)

(You can see the layout that we would like in a mockup, although slider that enables user to slide left and right to select a backup point and calender would be moved under breadcrumbs just above IconView. The backup mode itself would probably be activated from view menu.)

The first question that I have is what kind of view Nautilus team has on this integration? The most obvious benefit for this is that we can provide high level of integration with Deja Dup and provide some nice usability improvements for users but on the other side I am not quite sure how would you guys take integration with a single program.

Would you be willing to provide more extensive API? Should I patch Nautilus if user installs Deja Dup (any recommendations on how to best accomplish this)? Anyone tried developing Nautilus with Vala or do I need to use C for all development?

Maybe do you have any other suggestions on how to best approach this problem?

[0] https://launchpad.net/deja-dup
[1] http://dl.dropbox.com/u/1493539/past-rest-mockup.pdf




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]