1
5 Comments

Relying on 3rd party APIs in your app - possible repercussions

Hey folks, hope you're all staying safe in these uneasy times and still feel energized about your businesses!

Recently I decided to start developing apps for myself instead of my employers/clients, which I have been doing professionally for more than a decade, and see which ones get some traction/outreach.

One of the first ideas is a dictionary-like app I would happily use myself that I called QUIKen. (You can already check it out on app.quiken.me - but that's out of scope of this thread.) It's intended to be a go-to source of information about unfamiliar English words for non-native speakers such as myself - and I also have a myriad of ideas of how to make it better.

However, I don't plan and want to build a real dictionary. Instead, I picked an API of a dictionary I was using myself, Merriam Webster, and started using it for the prototype I've built. It works, but I am uncertain about a few things in regards to relying on 3rd party API assuming the app gets adapted by users and becomes commercially successful.

  • How do I go about API response caching? I am not against paying for API access but I wanted to optimize for a possible popularity in the future with pre-caching most popular words in my local DB. However, I saw Oxford Dictionary API terms mentions they prohibit caching which I didn't even think about - but it makes sense indeed.

  • Who owns the data after all? And what are the (legal) limits of what I could do with 3rd party data, either obtained via an API or scraping of any sort?

  • I suppose this particular project won't reach this phase but let's say it does. And it heavily relies on 3rd party data source for word definitions, pronunciations, etc. Even cached - but still gathered from an independent vendor. How safe is it to operate like that? What stops these "data owners" from just blocking my access or requesting predatory rates/whatsoever, which can be a death of my tiny enterprise?

Maybe something else I am missing in regards to using 3rd party data/APIs?
I'll be happy to hear your thoughts and grateful for sharing experiences.

Cheers!

on April 15, 2020
  1. 1

    I can't speak towards this legally as I am not a lawyer, but you should probably look into their terms of usage and go with that. My guess is that scraping is probably not allowed (again, check their TOS) either. To your last point, I think, unfortunately, there's nothing that stops them. But if you do get big and successful with this, you could definitely talk to the 'data owners' and see if you can come up with a specific agreement on how to use the data, etc.

    1. 1

      I read the ToC and it explicitly prohibits data replication and API use in commercial apps. It's interesting now how I could go about developing a useful and successful language-learning app without actually doing any of those 🤔

      1. 1

        Is there a way that you could get around calling the API and have the users crowd-source the words somehow? (probably not ideal, but ...)

        1. 1

          Oh well, in my particular case it's going to ruin the entire idea since I need quality translations for a dictionary-like app. However, the approach is a really good one for other cases, I suppose. Thank you!

    2. 1

      Hi Peter, yes, thanks - we perhaps share the same common sense here.
      But I am not sure is relying on 3rd party data too bounding in general or not.