rethink-project / dev-protostubs Goto Github PK

The error handler at:

https://github.com/reTHINK-project/dev-protostubs/blob/master/src/protostub/google/GoogleProtoStub.ps.js#L298

should update the protostubs status with status "disconnected":

            _this.sendStatus('disconnected', onError);

The Observer should add a Listener to prostostub status url which requires to send the protostub url in the response body to the initial request used to deploy the protostub:

https://github.com/reTHINK-project/dev-protostubs/blob/master/src/protostub/google/GoogleProtoStub.ps.js#L83-L85

Very similar to GoogleProtoStub, UserActivityObserver hyperty is responsible for loading the correct protostub, depending on the OS:

• Android: GoogleProtoStub (also for Windows?)
• iOS: StravaProtoStub

To avoid code duplication, a FitnessProtostub is to be created.

According to implementation feedback some changes were done in the specs:

The Configuration json object received as input by the P2P Requester constructor, has a new field remoteRuntimeURL containing the RuntimeURL of the remote P2P Handler Stub.

In addition, the P2P Handler Stub should also fire a new event as soon as it receives an invitation from a remote P2P Requester containing in the body.

Updated specs here.

When App goes to "sleep" mode, for some reason event bus handlers are lost and are not recovered.

Proposal: before messages are published on eb, check handlers are registered otherwise re-register them.

I did a little fix on dev-runtime-core (branch dev-p2p-interworking-merge).
commit

On your protostub when you are using access token, a little fix needs to be done
Your code

You need to use "msg.body.identity.access_token" instance of "msg.body.identity"

Is it ok for you?

To adapt data retrieved from Mobi.e Gateway and publish as a Vertx Stream.

Suddenly, GFIT is not working:

• Scope not included to read data for this type of source.: fixed by using https://www.googleapis.com/auth/fitness.activity.read and https://www.googleapis.com/auth/fitness.location.read as scopes

To adapt data retrieved from GIRA Gateway and publish as a Vertx Stream.

Check here on how to configure it and how to generate access tokens for google accounts.

Add the required end points here and use as a reference the slack authorisation definition.

Probably it will be required to include some access token OAUTH related functions into the OIDC lib

As a (better?) alternatibe OAUTH access token related function could be moved to a separared lib.

As discussed here

Some of the APIs used are not supported by the nodejs and should be replaced by something else

Users are complaining some activities are lost when there is some delay (> 1 day delay) to sync the device with the platform.

Try to replicate and check the API pooling policy

To be used to manage access to GIRA user data. Still no info about the protocol to be used

To be used to manage authorisations to access MObi.e user data. Should be compliant with auth 2.0

Check Google OIDC implementation

To be used to manage access authorisation to EDP Distribuição user data by using a proprietary interface. The main data flows should be:

1. get login page
2. post authentication data with CGUID
3. redirect with authentication / authorisation result + token based on CGUID

See spec here

Create a Google FIT protostub with support to Walk and Bike monitoring.

It should include a Context Reporter for walking and biking context data objects objects.

specs to be provided in a separate doc

• Verify that a Data Object Reporter created by the Protostub can invite a User Activity rating Vertx Hyperty which subscribes it.

• frequency of information retrieval (in config)

• query for "new" sessions (between start-end timestamps)

• timestamp of last session read (query sessions that ended after last session) (DO lastModified field)

• timestamp/date of when app was activated (only account for exercise that happened after that timestamp) in Data Object Storage (DO creation time)

• resume DO reporter

• avoid pop-up blocking

see spec

Based on current WebRTC hyperties. See:

• main arch
• detailed workflow

• complete state machine implementation. Catch network connectivity failures and call sendStatus() accordingly

• adapt to how vertx side failures are handled (see reTHINK-project/dev-java-hyperty#40).

• standardise vertx side successful responses where only the body.value is returned from Vertx

• remove DSM / Wallet specific logic eg currently public wallet reporters are created without being demanded

• when restarting the app, instead of just querying for user activity after some time, make an initial query to account for the time between app uses

https://api.slack.com/docs/presence-and-status

As required by T6.3 is necessary to implement a IWstub to interact with an IMS lab environment.

High level details fo the IW stub are documented in this presentation: GDocs

Only apply timeouts when connections are not ready or in attempts to re-connect

Currently delivery of messages have a delay of at least 2 sec:

https://github.com/reTHINK-project/dev-protostubs/blob/master/src/protostub/vertx_app_stub/VertxAppProtoStub.ps.js#L113-L144

Instantiate the Vertx EB in the constructor if possible

Currently it is processed as a new subscription invitation but it should be just forwarded to vertx runtime

https://github.com/reTHINK-project/dev-protostubs/blob/develop/src/protostub/vertx_app_stub/VertxAppProtoStub.ps.js#L217-L220

Each time there is a change in the wallet eg a new transaction, the Vertx Wallet Manager Hyperty should only publish in the Stream the new transaction, not all transactions. Currently, the amount of data published will continuously increase becoming difficult to handle.

In order to support this, the data format used should be similar to reTHINK UPDATE body message.

Integrate Slack Open API with a Group Chat Observer.

Ensure the token returned is well formatted here.

To adapt data retrieved from EDP Gateway and publish as a Vertx Stream.