- price of product
- food mass of product
Those vary invisibly all the time, and it'd be great to have a record of it.
After talking to people about it though, it became apparent that most grocery shoppers already have an idea of where they are going to buy groceries, making location more important than price. It's a razor thin industry, and even the price-sensitive coupon-using folks almost always just go to the closest Safeway equivalent.
So many food brands are now just meaningless words in someone's brand portfolio. Especially silly when the two main competitors for an item turn out to now be owned by the same parent and rationalised to be made in the same factory.
They get traded around like football cards.
We have them :-) We also store every single revision of the product, but we don't display the changes yet.
Join us on http://slack.openfoodfacts.org to see how we can just do that.
If your ingredients data is getting stale, you could reward users with coupons for stale-data items, so some users will take pictures of the ingredients list (kind of a 'bounty' list). You could do this for new items too - e.g. a popup after the picture is taken saying you can get $0.50 off if you take a picture of the UPC and ingredients list. Do a visual diff on the ingredients photo to see if the they have changed, and if so, OCR it to the database.
One is easy, the other is hard.
There are APIs and libraries for OCR, so that might not be so bad (even on a mobile device). A low-end phone is faster than the first PCs I did OCR on anyhow, and about 20 years more research has gone into algorithms.