aiyprojects-2018-04-13 'Check Cloud' provided URL gets a 404 'Not Found'

[juggledad]

I installed the aiyprojects-2018-04-13.img on a RPI3. When I run the 'Check Cloud' script it says:

Please follow these instructions to get Google Cloud credentials:
https://cloud.google.com/speech/docs/getting-started#set_up_your_project
and save them to /home/pi/cloud_speech.json

however that URL produces a 404 error

In addition, the https://aiyprojects.withgoogle.com/voice-v1/#users-guide step 11 says:

Find the JSON file you just downloaded (client_secrets_XXXX.json) and rename it to assistant.json. Then move it to /home/pi/assistant.json

while check_cloud.py says to name it 'cloud_speech.json'

This does not make it easy for someone to actually use this kit.

[drigz]

Hey @juggledad, thanks for the feedback! You only need to run Check Cloud when you want to use the Cloud Speech API (following the instructions here, which explain how to create cloud_speech.json). If you're using the Google Assistant (the default, which is described in the Users Guide) you don't need to worry about Check Cloud.

I'll leave this issue open to track the broken link, which we will fix in the next image.

[juggledad]

Well you still need to fix the difference is which name to call the json file.

In addition, I click the 'Start dev terminal' and the V1 guide says to enter:

src/examples/voice-v1/assistant_library_demo.py

However there is no voice-v1 folder and if I use

src/examples/voice/assistant_library_demo.py
it fails with:

pi@voice:~/AIY-projects-python $ src/examples/voice/assistant_library_demo.py
Traceback (most recent call last):
  File "src/examples/voice/assistant_library_demo.py", line 73, in <module>
    main()
  File "src/examples/voice/assistant_library_demo.py", line 67, in main
    with Assistant(credentials) as assistant:
  File "/opt/aiy/projects-python/src/aiy/assistant/library.py", line 33, in __init__
    self._model_id = device_helpers.register_model_id(credentials)
  File "/opt/aiy/projects-python/src/aiy/assistant/device_helpers.py", line 93, in register_model_id
    r.raise_for_status()
  File "/usr/local/lib/python3.5/dist-packages/requests/models.py", line 935, in raise_for_status
    raise HTTPError(http_error_msg, response=self)
requests.exceptions.HTTPError: 403 Client Error: Forbidden for url: https://embeddedassistant.googleapis.com/v1alpha2/projects/voice-1-201708/deviceModels/voice-1-201708-voice-kit

Did anyone test the latest release with a V1 kit?

[drigz]

@juggledad What difference in filenames are you referring to? For the Assistant you need to create assistant.json and for Cloud Speech (& Check Cloud) you need to create cloud_speech.json.

Thanks for pointing out the voice-v1 mistake - we'll fix that in the next update.

The "Forbidden for url" error looks like #321 - if you re-enable the Assistant API does it help?

[juggledad]

@juggledad What difference in filenames are you referring to? For the Assistant you need to create assistant.json and for Cloud Speech (& Check Cloud) you need to create cloud_speech.json.

Why would you have two different names for the credientials file?

Ok, re-enabling the Assistant API got it to work

[drigz]

Why would you have two different names for the credientials file?

They use different kinds of credentials. The Cloud Speech API uses a Service Account (cloud_speech.json contains a private key for the Service Account) whereas the Assistant needs access to your user account (assistant.json contains the "client secrets" used to start the process of getting partial access to your user account). Sorry if that's a confusing explanation - the authentication side of things is quite complex.