by dandv (Posted Fri Jan 13, 2017 9:31 pm)
You lose users by having unnecessary complexity. How many users genuinely need a difference between "categories" and "keywords"? I consider myself a power user (worked as a software eng. at Google and Yahoo!, have been using image management software since 1998, and have a keen attention to detail) but never needed this distinction. I also use the most powerful photo management app for Android (F-Spot) and its tagging capabilities are excellent while using only one concept, tags. F-Stop does use a database of tags, but only for performance. It stores the metadata within the images, to avoid lock-in. This is the best of both worlds.
My strong suggestion is to simplify tagging by unifying keywords and categories in the UI, and storing the same underlying metadata in all applicable formats (e.g. both XMP and IPTC). That will ensure compatibility with tools that support either format (e.g. the most popular image viewer on Windows, Irfan, only supports IPTC, while gThumb, F-Stop, Google's Snapseed and many others use XMP). It will also avoid the perplexing situation I'm illustrating in the screencast below (click it to see the animation).
Read Main Topic
m.Th. wrote:Do we loose [sic] users more than we gain if we change „categories” to „keywords”?
You lose users by having unnecessary complexity. How many users genuinely need a difference between "categories" and "keywords"? I consider myself a power user (worked as a software eng. at Google and Yahoo!, have been using image management software since 1998, and have a keen attention to detail) but never needed this distinction. I also use the most powerful photo management app for Android (F-Spot) and its tagging capabilities are excellent while using only one concept, tags. F-Stop does use a database of tags, but only for performance. It stores the metadata within the images, to avoid lock-in. This is the best of both worlds.
My strong suggestion is to simplify tagging by unifying keywords and categories in the UI, and storing the same underlying metadata in all applicable formats (e.g. both XMP and IPTC). That will ensure compatibility with tools that support either format (e.g. the most popular image viewer on Windows, Irfan, only supports IPTC, while gThumb, F-Stop, Google's Snapseed and many others use XMP). It will also avoid the perplexing situation I'm illustrating in the screencast below (click it to see the animation).
Read Main Topic