I’m currently evaluating graph.cool and have a couple of questions: 1. when will “advanced” permissions be available? (e.g. authors can delete only own posts / admins can everything) 2. I can export a projects schema - but cannot import it into another project!? 3. filter/search in json fields not possible? 4. filter/search in multi-value fields not possible? 5. subscriptions fires only for top-level mutations (not for nested/relation data changes), so I would need seperated subs for each sub-type!? Thanks 🙂

@sokki thanks for your questions! I'll answer them one by one: 1. pretty soon 😉 We've run this feature successfully in our beta program and are finishing up the API. Here's a webinar to see how it works: 2. You can clone a project by expanding the project list and clicking the duplication button to the right of the project name. 3. That's correct! We're researching different approaches currently. You can follow this feature request for updates: https://github.com/graphcool/feature-requests/issues/148 4. If you are referring to list fields, that's also correct. Here's the corresponding feature request: https://github.com/graphcool/feature-requests/issues/60 A current workaround is promoting the scalar list to a "to-many" relation to a full-blown model instead. 5. In general, that's right, but I'm interested to hear your specific use case because using nested mutations you can get around this depending on the use case. (Feature request: https://github.com/graphcool/feature-requests/issues/146) Hope that helps! 🙂

Thanks for the quick answers @nilan - awesome 👍 1. great 🙂 2. unfortunately, this won’t work in our scenario (multiple developers in multiple environments). Also, I couldn’t “merge” changes back. I guess, the upcoming editable schema will help here. 5. quick example would be:

order = {
  basket: {
    products: [
      {
        features: [
          {
            name: 'color',
            options: ['red']
          }
        ]
      }
    ]
  }
}

All those informations are gathered in parallel by multiple bots and humans. I guess, the need for real-time in this scenario is not a common case, but crucial for us 🙂 I’m not yet deep into graphQL, but this would already be possible by dynamically creating seperate subscriptions for each-subtype!?

2. exactly! currently, the clone project feature is probably the best way to handle multi-environment setups. We're currently working on improvements regarding the edit schema feature and also a CLI for schema migrations 🙂 5. Yes, you can subscribe to the different subtypes each. It will be a bit tricky to combine all the results together, but it's possible 🙂 I'm happy to answer more questions about Graphcool or GraphQL, thanks for your interest!