2

I currently have 30gb of documents and another 20gb of images. I am due to start college soon, where I will pick up even more.

I would ask that Dolphin be given the ability to auto-file files. Examples are included below;


The documents folder has 30 gigabytes of odt, pdf and other such data avaible. Using the meta-data about these files, Dolphin could file/filter them by author/time of creation/time first on the system/etc

Likewise, the Music file could do a similar feat with album collections, file them away by author (and even making the icon the cover-art?)

Photos could be dealt with the same way.


I realize this may not sound like the best of ideas, with Strigi fully operational and all, but with a collection of files this large its nessercery to dive right into them and pull out what you want.

Mind you, once NEPOMUK (and possibly Scribo) get online and fully operational, I doubt this'll be as needed as it is now.

So, anyone else want Dolphin to do their filing for them?

I think the idea of having the system assist in filing is good but it will not be to everyones tastes. I think a lot of users find organising files hard. So they will either love not having to think about it or resent having a system muck with their careful design.

iTunes can "organise" your music files for you for example. Some love it, some hate it. Either way, iTunes having the ability to "organise" files for you based on metadata does make you want to get the metadata right. So this might actually have a *benefit* for strigi/nepomuk. Carefully integrated I think this would be a great feature.

One thing I would ask is can this be put up the tree a bit? When I save/create a file in a KDE application why not allow the user to set some meta-data tags and then provide the option "Allow KDE to manage this file" which would automatically generate a path and filename based on the metadata. File names and paths are just metadata after all.

Agreed. I cannot see this being of use to those who DO carefully organize and keep their files, as well as those who have a smaller set of files to fuss over.

I think krename can do what you are asking for here.

I don't see how a batch renamer can file away documents, music and photos into an orangized system..

It can use file properties like file type, creation/access/modification date, file name patterns, and audio and picture tags to rename and/or move files. Check out the "plugins" section to see what sort of functions are available, and use the "filename" tab to move them based on those rules.

Hmmm...it seems the repo only stocks an older one.

Hey, is there any chance that can be integrated into file saving (AKA my suggestion above) so that entering useful metadata might become the "main game" and users have the option of leaving file naming up to the DE? This is quite a distinct idea to batch renaming (bulk filing) which is krenemes purpose. But obviously one that has a lot of components in common.

Then there is no need for "re"naming Is this worth submitting as a separate idea?

Dante, are you talking about something like GNOME's Zeitgeist?

In a way; though my problem is a bit more complex;

Having loads of files means I have to go and file everything away manually. One album, for instance, is filed like this;

Home-> Music -> Doctor Who Music -> (subfolders) Series 1, Series 2, Series 3, Series 4.

Likewise, I'd like the system to do that, but automatically for all files.

GNOME's Zietergiest seems very interesting, though seems to be a search engine more then a file system filter.

Something like in Amarok Auto Organize files?

Could this be done with the "magic folder" plasmoid? I have not used it very much myself but I know one can create pretty complex filters and then just drop your files there.

As I see it this idea is completely different to "finding" files (as implied by filtering). It is much more like "auto organise" files.

Auto organise is common for music files and is one of the reasons for getting the metadata right for music files.

If auto-organise was available as a save option for all files there would be a greater inclination to "get the metadata right" for more files. Which may in turn help desktop search work more effectively.

I don't see this as an extra burden BTW. Instead of naming a file explicitly you assign it some tags instead. The system comes up with a name based on these tags and other metadata from the file (title etc).

If I'm looking for a file, great! I can get to it.

But if I'm looking for a specific type of file; a group, rather then a singular document...then we have a problem.

I'll put this up on the wishlist.