- We currently support fetching code with username = 'oauth2' and
password = <access_token>.
- Trying to _push_ code with the same credentials fails with an authentication
error.
- There's no reason this shouldn't be enabled, especially since we allow the
OAuth client to create deploy keys with push access:
https://docs.gitlab.com/ce/api/deploy_keys.html#add-deploy-key
Gitlab::Auth.find_with_user_password is currently used in these places:
- resource_owner_from_credentials in config/initializers/doorkeeper.rb,
which is used for the OAuth Resource Owner Password Credentials flow
- the /session API call in lib/api/session.rb, which is used to reveal
the user's current authentication_token
In both cases users should only be authenticated if they're in the
active state.
- cleanup formating in haml
- clarify time window is in seconds
- cleanup straneous chunks in db/schema
- rename count_uniqe_ips to update_and_return_ips_count
- other
We accept half a dozen different authentication mechanisms for
Git over HTTP. Fairly high in the list we were checking user
password, which would also query LDAP. In the case of LFS,
OAuth tokens or personal access tokens, we were unnecessarily
hitting LDAP when the authentication will not succeed. This
was causing some LDAP/AD systems to lock the account. Now,
user password authentication is the last mechanism tried since
it's the most expensive.
- The `scopes_form` partial can be used in the `admin/applications` view
as well
- Don't allow partials to access instance variables directly. Instead, pass
in the instance variables as local variables, and use `local_assigns.fetch`
to assert that the variables are passed in as expected.
- Change a few instances of `render :partial` to `render`
- Remove an instance of `required: false` in a view, since this is the default
- Inline many instances of a local variable (`ip = 'ip'`) in `auth_spec`