These are chat archives for Ruhrpottpatriot/GW2.NET

30th
Dec 2015
Robert Logiewa
@Ruhrpottpatriot
Dec 30 2015 13:58
package managemnt is actually better now, just different. I think you add the namespace you need in the project.json and the rest is done automatically.
Steven Liekens
@StevenLiekens
Dec 30 2015 13:59
yeah I figured it out
Robert Logiewa
@Ruhrpottpatriot
Dec 30 2015 13:59
As for why I don't just do a Take(200)? Think of it: What If the user requests 202 items from the API? Then we'd do two requests where one is 198 items too large
my algorithm, which is really a two liner, just calculates the page sizes and give them back as an array
thus we only request the items we need
btw:
The items endpoint, does not have an ids=all parameter
so we need to do it different
Steven Liekens
@StevenLiekens
Dec 30 2015 14:00
with paging
Robert Logiewa
@Ruhrpottpatriot
Dec 30 2015 14:01
yep
Steven Liekens
@StevenLiekens
Dec 30 2015 14:01
ids=all is useless
Robert Logiewa
@Ruhrpottpatriot
Dec 30 2015 14:01
i know
but it'd make querying easier for a FindAll()
Steven Liekens
@StevenLiekens
Dec 30 2015 14:01
unless they ever add endpoints that support ids=all with more than 200 results, let's just stick with page requests
Robert Logiewa
@Ruhrpottpatriot
Dec 30 2015 14:01
I propose:
  1. do a discover
  2. Pass all the ids to a suitable Find(identifiers) method and return
that's the easiest way
Actually, if the request is not a single item request I'm only using paging right now
Converters did change a bit, I'm experimenting a bit still, but I'm trying to make the converters type safe, but non generic so we can pass them around easier.
The goal is, that you only need to pass only the necessary converters, and not do stuff like pass the whole batch converters etc.
Higher order is a wunderful thing :P
Robert Logiewa
@Ruhrpottpatriot
Dec 30 2015 14:08
I'd like partial function application too, but hey, C# not totally functional... I'd have to go with F# for that :)