General information

Since we are not only trying to take into account the wishes that have come together at ideas.runalyze.com, but also technically to make the API a bit cleaner and at the same time keep it backwards compatible, the programming, documentation and especially the testing takes significantly longer.

This concerns the "Personal API" as well as the Third-Party API.

The following topics will be possible as soon as the programming is finished and the refactoring has gone live:

Personal tokens can be "scoped" and do not have access to all endpoints immediately.
Third party apps can also be scoped more accurately and can get more access. Among other things, also read.
HRV values can be saved with the different units (important for e.g. the iOS app Healthfit)
Missing endpoints around Daily Notes, blood glucose, body temperature will be included

In a later step these wishes will be implemented:

Extend Personal API to allow to set tags, equipment, sport type&subtype during upload
https://ideas.runalyze.com/posts/28/extend-personal-api-to-allow-to-set-tags-equipment-sport-typeandsubtype-during-upload
Reading activities (including data streams). This point will take some time, because an activity includes a lot of data.

Status Update

Status: 2023-03-21
Still work in progress.
Status: 2023-02-26
Because of sickness, a lot of support requests and coding other features currently no further update.
Status: 2023-02-08
Because of sickness no further updates on this. The progress on replacing the old third party library are already almost done and just need fine tuning.
Status: 2023-01-22
One big change what is missing is replace an old library we are still using. As that library is the main part of the authentication of the third party library this is quite a big task as we will replace some parts of the replacement with a custom solution we need.
Status: 2023-01-20

Currently Michael is testing through the different endpoints, verifying that before/after the API works the same. So far it looks already quite good, but we need to test everything completely as there will be no point of return after the deployment.
Especially the documentation of the endpoints (even if relatively automatically created) has to be gone through.