Well the camera source is going to be opened up so we heard, so this may well become a useful thing to do (along with lots of other cool camera processing stuff .

The only problem I envisage is that the decoding takes processing, and if you're not looking for barcodes it is a major waste of cpu cycles/power to still look for them in every frame.

With that said, I guess the barcode processor could be enabled per-session, in the same way I envisage a smile/face/blink detection plugin to be enabled per-session.

Would people prefer this type of setup to the monolithic one I have at the moment? Certainly until the camera source is opened, I could write a very basic gui (not so hot on the guis you may have noticed to scan and then output the decoded data over dbus. Will anyone use such a thing?