Well, yeah. But... I don't like the random character strings in the filenames of my images. I would rename them, as I stated, but clearly, the database wouldn't appreciate that and would throw a fit. There should atleast be an option for you to rename files--even if they don't share the same name as the image's title.
I would avoid any link between the filename and the title whatsoever. It should only be used if nothing else was available.
Preferred naming of a picture:
Automatically import it from the corresponding IPTC field, but allow it to be edited.
Add title manually at the time of import.
Use filename
But import from IPTC should always be the preferred.
Oceanwatcher, that's the idea. Once we have EXIF/IPTC support, then it'll be easy, but I'm waiting to mirror that in the database so I don't have to go to the file each time, and that requires quite a bit of modification to the database. I think it's scheduled for 1.1
Sounds right to do it that way.
But remember to leave in a function to re-read the IPTC data from the files if the database gets messed up at any point.
And maybe that function should have an option of conflict solving if differences were detected... Like side-by-side comparison and ability to choose the one you want.
Can you give me a ballpark timeframe/guesstimate on when we can see this version out?
Also, it could be nice to be able to start building the gallery now, and then just hit a button in 1.1 to get correct Title and caption etc. when it is ready.
Like a "reload titles from image metadata" link? Yeah, I could see that.
Times are hard to estimate, but let's see.. I'm graduating on the 21st... after that I'm working, but should be able to code nights and weekends sometimes, and I really want this metadata support in the 1.1 version (with sub-albums) so let's give it six to eight weeks, conservatively. I could be done a lot sooner if I find the time.
The technical stuff shouldn't be too hard... I'll mirror the file metadata in the database (much much faster) and probably use the mtime as a last-modified date to see if I need to reload it. And of course, if the database is ever corrupted, it will know to reload all metadata from the file again, just as if the file were new. Sounds pretty straightforward.