These are chat archives for Ruhrpottpatriot/GW2.NET

Dec 2015
Steven Liekens
Dec 19 2015 10:13
Discover() is useful if you keep a local copy of everything in the API and you want to check if your copy is complete
think when new items or recipes are discovered and added to the API
you probably only want to get data for those newly added IDs and skip the IDs that you already have
Steven Liekens
Dec 19 2015 10:20
I think the intended workflow goes like this:
  • first run: get everything, using page bulk requests
  • daily / hourly run: get new IDs only, using ids bulk requests
  • after a game patch (new build number): same as first run
there's also ?ids=all, which I don't think is very useful
?ids=all is just shorthand for ?page=0&page_size=200
but only for endpoints that have <= 200 IDs
Steven Liekens
Dec 19 2015 10:26
ids=all will return an error for endpoints with more than 200 IDs while the page request does not
if you want to drop APIs, drop ids=all support
Robert Logiewa
Dec 19 2015 12:59
That makes sense. In the new repository layout I'm aiming how to reduce the workload the end user has to do (while keeping customisation options as much as possible).
Currently looking into the observer pattern
Steven Liekens
Dec 19 2015 13:19
observer doesn't work for pull based apis
you'd have to implement polling