I have been looking for a gallery that has certain features, and I have realized that they just don't all exist in the same place. I have enjoyed Gallery2 for a while though so I figured with a project of any scale I should focus on building it in the next version.
I have been looking for a standard way to include image region annotation (like most of the social sharing sites now support) that can be placed into the head of a jpg (in an exif field).
Currently I have found just a handful of projects doing this, nothing standard. Microsoft people tags has the most promise to become a standard, and even those are clunky. Then there is Fotonotes, which is more a script kit and proof of concept than it is a well documented living standard. So where does that leave us (anyone who wants to have image annotation)?
My thought is to expand and alter slightly the formatting of an html4 style image map. Include the map as a new exif tag. Then create javascript and/or php script to take the data to make annotations visiable to a viewer. Then make tools that allow easy annotation through javascript while viewing.
This will be a very long term project, as a father of young twin girls I have a lot of things that will get in the way. However I am sick of waiting for other people to do it. What I need are design considerations here... on the front end. I have lots of design background, but i will have to learn most of the web methods and languages almost completely from scratch.
Things I have come to that I don't yet have ANY answer for... much less a good answer
*when performing editing with javascript, how is a save securely processed?
*when processing the exif data, does it make sense to do this on the back end(server side) and cache it per image, to maximize compatibility with different browsers, or does the added flexibility of real time creation where the parser knows more about the display environment better (client side)?
*how granular should scripting like this be? Is the goal maximum use (like lightbox and other JS scripts) or is the proven design practice that tighter, more directly coupled solutions work better, and are just more work to set up?
most of these probably just show what I already stated about not being that familar with web based programming, but any help, direction, suggestions etc, would be welcome!
PS. As I make some progress I would be eager to accept assistance from someone with more background in web development.
Posts: 22890
Just a link to an existing module that does something similar:
http://codex.gallery2.org/Gallery3:Modules:tagfaces
Saving data in the EXIF header will be troublesome as you will have to find a existing 3rd party image processing library to do that and create a module in G3 to do that first. I would store the data in a db; per item. It will be a lot easier as a first step.
There is a few JS libraries that do tagging/maping. I would use some that use the jquery library since it is already loaded for the some of the JS on the default theme.
Good luck.
Dave
_____________________________________________
Blog & G2 || floridave - Gallery Team