NASA API APOD as Falcor routes about apod-graphistry-falcor-routes HOT 4 CLOSED

@mattiaerre in case you're interested, we've finished building in end-to-end streaming (both "live queries" and "subscriptions", in graph-ql parlance) into our @graphistry falcor repos, even maintaining batching/request de-duping for streaming requests. Docs are still sparse, but the unit tests are there. Falcor-react-schema provides react integration and dynamic compositional route definitions, which we've been running a in production for about 6 months without issue. LMK if you have questions, or if I can help in other ways.

from apod-graphistry-falcor-routes.

thank you very much @trxcllnt I will definitely try your libs; so basically I should be just able to install them and change my imports and everything should just work right? there is anything else, in particular, that you recommend having a look at?

from apod-graphistry-falcor-routes.

@mattiaerre the only backwards-incompatible changes are removing path-syntax and HTTPDataSource from the core falcor project, for speed and bundle size (currently @ 20kb gzipped!). The rest should be bug fixes, performance improvements, and new features. If bundle size isn't a concern, you can import @graphistry/falcor-model-rxjs, which wraps the public methods as RxJS5 Observables, and adds path-syntax support back in.

We also rely on webpack's DefinePlugin to rewrite source identifiers to our internal keys, so if you're doing any webpack aliasing, be sure to reference the dist files. The falcor.all build bundles in the dependencies from falcor-path-utils, and the falcor.node build has them externalized as regular CommonJS requires.

We've also made changes to the falcor-router to support streaming (which can be used in conjunction with the sinks/sources in falcor-socket-datasource).

The most visible change is that all path values and invalidations will be communicated back to the client, and the client cache will update accordingly. Previously, the client would ignore any extra data the server decided to transmit, and would also ignore invalidations returned by a get or set.

So if your routes communicate more data than requested (e.g. the client asks for path/foo, but the route emits path/foo and path/bar) or you have get/set routes that return path invalidations, the client cache will update to reflect this, full-stop.

from apod-graphistry-falcor-routes.

@trxcllnt I am going to use your fork from now on; I just updated and published to npm this component (and also changed its name). v1.1.0 is going to be used by supersonic-lobster that I am going also to update to the graphistry fork

from apod-graphistry-falcor-routes.