• 0 Posts
  • 23 Comments
Joined 3 years ago
cake
Cake day: June 1st, 2023

help-circle
  • “Rendering my cash invalid with no recourse” might be on the extreme end of this, if the places that its checked are ATM and banks - so long as it’s still exchangeable for goods/services (e.g. not being checked by stores)

    I’ll keep an eye out for kore info though.

    But yes, with serials are tracked when taken from/put in ATM/bank, it is not anonymous. Potentially breaking larger bills at common places (stores/gas stations) could mitigate this, but theres still information about where you’ve been inherent in that.









  • Paperless-ngx does include OCR, as well as supporting document types (which can have fields, etc) - but there is no built-in way to intelligently extract field values. You can use the python API to access & update the data and fields. So field extraction via your own code is feasible.

    Given the large variety in receipt layout & potential for habdwritten totals after tips - I’d encourage part of your workflow to include manual ispection/correction of every processed receipt - or at the bare minimum that you include check-in points where you verify your end balances 100% match after all transactions have been entered, so you can detect & root-cause errors ASAP.





  • Another interesting one I’ve seen: Upside.

    It isn’t necessarily encouraging to buy a specific product, but rather to buy from specific businesses.

    To the consumer: they look at the app and buy gas from participating stores, which typically show as having a 5-10 cent discount on gas. The customer uploads a picture of their receipt to confirm the purchase and amount, and get credited that discount. Theres a monthly purvhase target that unlocks a 1 time discount of 30 cents per gallon, etc.

    To the business: they pay a fee to Upside to be a participant. Upside promises to only participate with a certain % of stores in the area. It’s likely the business also pays for some of the discounts obtained.

    Essentially paying upside is meant to increase sales volume. While a gas station could just dicount their gas, they are paying Upaide to effectively advertise and steer consumers to THAT store. As Upside is national, many custpmers use it, even in unfamiliar areas to choose gas stations.

    I say not a specific product, because my understanding is it’s extended to other business categories besides gas.

    Market manipulation, ultimately. I’m sure in some part they are sellong your location & purchase data as well.








  • This looks awesome, and I’ll explore it a bit. I love this being open source, and it looks very clean.

    I’m going to pull a prior post, because I think ALL could be potential improvements for your utility. Feedback mainly sensible for Kana, but it could be applied to kanji with significant effort.

    I like these resources for practicing kana:

    • https://gohoneko.neocities.org/learn/kana
      • it’s like most other “type what you see” quizzes, but I like that it has a good selection of handwritten fonts that normalize different strokes/styles, so you can practice identifying the character written a number of ways
    • https://studykana.com/practice-reading
      • just a typing test. Gives you an excuse/motivation to try to go fast. I wish it had more words, or a way to upload a word dictionary, etc. (I think your app already includes a time test but I havent yet explored. I like that the above link isn’t multiple choice)

    I would LOVE for it to detect when you commonly confuse two characters and then offer to give you a short drill of just those characters to reinforce.

    OR if you could have a “good probability” of including easily confused characters in the multiple choice. Me/nu, wa/re/ne, chi/sa, can be easily confused. I havent done drills in a long while and I know roughly the shape I’m looking for, but would stuggle to differentiate some of these cases. With 3 multiple choice - odds are good I can guess whichever is present.